Rewriting the ROM
Andrew Huey
possible for volunteers to band together and rewrite all the features of
the calculator themselves? This way, the entire community could participate
in bug-fixing, enhancements, customizations, etc. I think we could do
something similar to what OpenBeOS is doing to BeOS (www.openbeos.org). In
fact, shouldn't it be easier, because the calculator, however complex it
is, is much simpler than a full-fledged, multitasking OS?
I've been an HP fan for years, and just bought my HP49 two days ago, so I
wouldn't know enough to contribute to such a project (yet). However, this
could keep the development alive. By the way, HP is still manufacturing
HP49's, even if they're not currently developing anything else, right?
I know I'm wrong, so shut me up and put me in my place.
--Andrew Huey
Micah
with a 4mHz processor. What would be really kewl is to write the HP ROM for
the TI calculator. You would just have to get over the keyboard issue.
This way, since TI still has calculator development ongoing you would
probably be able to port it to a new calc platform. How kewl would it be to
have the 10mHz processor with RPN and all the features of HP? The HP could
live on in its reincarnate version.
Micah
"Andrew Huey" <andrew...@attbiALL.comCAPS> wrote in message
news:ei776a...@127.0.0.1...
George Tsiros
> At this point it would be stupid to rewrite a ROM for a piece of hardware
> with a 4mHz processor. What would be really kewl is to write the HP ROM for
you're trolling, but you are doing it so subtly that not even you
realize it.
Why?
This "piece of hardware with a 4mHz processor" happens to be one of
the best calculators ever produced and a creation unlike any other you
will probably ever see.
That is one. The rest find it yourself.
Troll.
Steen Schmidt
> realize it.
Why is he trolling? He's right.
> Why?
>
> This "piece of hardware with a 4mHz processor" happens to be one of
> the best calculators ever produced and a creation unlike any other you
> will probably ever see.
That 4MHz piece of hardware is ancient, and was (hopefully) merely a
stepping stone for a better calculator. The HP49G is not much more than
HP48GX+www.hpcalc.org.
The '49 is slower than slow for most stuff a normal student will ever do,
and yet it was sold as a school calculator. The pathetic documentation at
release, the lack of HP backing, the many bugs the first 18 months and the
lack of new hardware made the HP49G a failure standing alone. One of the
biggest mistakes was to make it compete on the same market as the TI89/92
series. How could it?
Be sure of that many students, who before didn't know HP as a calculator
brand, now dislikes HP. HP have failed them in some way another - at an exam
or in a class. HP was so eager to apply to their promise of 1 MB free flash
ROM for the user, that the ACO wasn't allowed to use any more for the OS.
What about all the otehr promises on the boxes? Practically none can be
defended.
Step-by-step? Non-existent. TI-like SOLVER. Not by a longshot.
The OS and CAS was from the start very inconsistent, and that lead to
software impossible to fix bugs in. Look at it now - there are still newly
introduced bugs - there are even recognized bugs that the ACO couldn't fix,
since they were so embedded into the system.
I have written alot of software, and the most important things in that
process are specification, documentation, modularization and tests.
No piece of software is better than its design specification. When software
is not documented, you can't fix bugs, and noone else can work on it after
you quit or get fired. If software is not coded in highly specific modules,
you can't write a good test specification, and hence can't test it properly.
If you can't, won't or don't have the time for testing, the software is a
failure, since it'll most definetely contain bugs that'll take a long time
to find and be near impossible to fix. When old bugs are left in the code,
while newer ones are fixed, you risk introducing secondary bugs again. This
cycle will kill your project.
I'm sorry to say this, but from a professional perspective, the HP49G seemed
like a hobby project.
I won't speculate on the reasons for this, and this is not an attack on any
of the guys we know from the ACO.
It's merely a statement that I hope will open your eyes. The HP49G is not
fast. It's really not implementing anything new (the new that is is
practically unusable, because of the poor performance). It's not stable with
release ROMs (v1.18). It can't be fixed now. The things learned from it will
probably not ever be used in any way by HP. It's a failure.
In hindsight, the ACO didn't seem stable either. How can such an important
product be rushed like this? It was doomed to fail - HP didn't gain
anyhting, but HP is to blame for that. There was no support, and hence HP
didn't want this to succeed.
Regards
Steen
Micah
fund a project (well, aparently HP) that was at least 10 years outdated.
Don't get me wrong, I love HP calculators but they are behind the times.
For the same price you can get much faster hardware. 233 mHz isn't unheard
of in handhelds. And no comments about batteries, we have rechargeables.
Now, if HP hadn't killed the Expander that would have given TI something to
worry about.
And let me just pick on the 49g for a second. The keyboard is horrible.
This is my second that has keys wearing out. It can't do all of the things
with graphs that TI has implemented. Try to find the slope of the tangent
to a conic section.
It's a useless debate. If you want to spend thousands of hours to develop a
ROM yourself. Go for it. How much longer can HP still manufacture the
calculator line? Being optimistic, I can see all HP calcs being collector
items in (at most) 5 years. Where do you think TI will be by then?
Micah
"George Tsiros" <gts...@yahoo.com> wrote in message
news:df4089c6.02030...@posting.google.com...
Parisse Bernard
it is not objective anymore. Sure we had not enough time to
release a stable ROM at the beginning (but the current 1.19-6 release
is fairly stable), a faster processor would have been greatly
appreciated as well as inclusion of the documentation available
on the web page of Renee De Graeve. The HP49 is however a nice
calc, and we have introduced some concepts that are now imitated
by TI or TI fans like step-by-step (which exists despite your comment)
or EQW. And the fact that people continue to ask whether they should buy
a TI or an HP is a sign that it is in the competition.
I don't think however that rewriting a ROM for the HP49 (or any
calculator)
is a good idea because the architecture of the 49 was designed for
systems
with very low memory capacities unlike current hardware. I believe
that handhelds with e.g. linux OS have a much better future (and
that's not against a calculator with a keyboard: there are still
keyboards for PDAs).
Nick Karagiaouroglou
> > you're trolling, but you are doing it so subtly that not even you
> > realize it.
>
> Why is he trolling? He's right.
>
> > Why?
> >
> > This "piece of hardware with a 4mHz processor" happens to be one of
> > the best calculators ever produced and a creation unlike any other you
> > will probably ever see.
>
> That 4MHz piece of hardware is ancient, and was (hopefully) merely a
> stepping stone for a better calculator. The HP49G is not much more than
> HP48GX+www.hpcalc.org.
Never considered it this way, but now that you said it...
> The '49 is slower than slow for most stuff a normal student will ever do,
> and yet it was sold as a school calculator. The pathetic documentation at
> release, the lack of HP backing, the many bugs the first 18 months and the
> lack of new hardware made the HP49G a failure standing alone. One of the
> biggest mistakes was to make it compete on the same market as the TI89/92
> series. How could it?
>
> Be sure of that many students, who before didn't know HP as a calculator
> brand, now dislikes HP. HP have failed them in some way another - at an exam
> or in a class. HP was so eager to apply to their promise of 1 MB free flash
> ROM for the user, that the ACO wasn't allowed to use any more for the OS.
> What about all the otehr promises on the boxes? Practically none can be
> defended.
>
> Step-by-step? Non-existent. TI-like SOLVER. Not by a longshot.
Well Steen, Step-by-Step is "somehow" existent, but it this "somehow"
is not understandable by most of the people for using it. I think the
makers assumed that *everyone* that uses the machine will also have
the computer/algorithm knowledge that is necessary to interpret the
starnge messages of the Step-by-Step feature.
> The OS and CAS was from the start very inconsistent, and that lead to
> software impossible to fix bugs in. Look at it now - there are still newly
> introduced bugs - there are even recognized bugs that the ACO couldn't fix,
> since they were so embedded into the system.
Yeap, exactly. After a long (and very excited) time of using it, I get
the feeling that in the HP49G we have a battle of the OS vs. the CAS.
Those two contradict each other and fight for the right to take
control. When two elephants are fighting, the victims are the ants
(we), as people in India say. (Bhuvanesh ? ;-) )
> I have written alot of software, and the most important things in that
> process are specification, documentation, modularization and tests.
>
> No piece of software is better than its design specification. When software
> is not documented, you can't fix bugs, and noone else can work on it after
> you quit or get fired. If software is not coded in highly specific modules,
> you can't write a good test specification, and hence can't test it properly.
> If you can't, won't or don't have the time for testing, the software is a
> failure, since it'll most definetely contain bugs that'll take a long time
> to find and be near impossible to fix. When old bugs are left in the code,
> while newer ones are fixed, you risk introducing secondary bugs again. This
> cycle will kill your project.
And so it goes to a zombie project. Killed but undead. (Perhaps that's
why I still like it ;-) )
> I'm sorry to say this, but from a professional perspective, the HP49G seemed
> like a hobby project.
>
> I won't speculate on the reasons for this, and this is not an attack on any
> of the guys we know from the ACO.
>
> It's merely a statement that I hope will open your eyes. The HP49G is not
> fast. It's really not implementing anything new (the new that is is
> practically unusable, because of the poor performance). It's not stable with
> release ROMs (v1.18). It can't be fixed now. The things learned from it will
> probably not ever be used in any way by HP. It's a failure.
Oops! I wouldn't go that far to say that it is a complete failure
Steen. You can do many things with it, unbelievable things. But only
if you had time to experience its strange behavior, which of course
can't be assummed always.
> In hindsight, the ACO didn't seem stable either. How can such an important
> product be rushed like this? It was doomed to fail - HP didn't gain
> anyhting, but HP is to blame for that. There was no support, and hence HP
> didn't want this to succeed.
Now Steen, you mean they (HP) torpedoed the project themselves right
from the start? Why should they do that?
Greetings,
Nick.
Steen Schmidt
> Oops! I wouldn't go that far to say that it is a complete failure
> Steen.
A failed product (couldn't sell on the proposed market), not that much in
the sense of a failed calculator.
> > There was no support, and hence HP
> > didn't want this to succeed.
> Now Steen, you mean they (HP) torpedoed the project themselves right
> from the start? Why should they do that?
I don't know why, but they did. The ACO didn't get any funding (economical
as well as knowledgewise). Many of the promising projects were abandoned
without obvious reason. The ACO was closed, presumably because they didn't
make any money - it was a *development* departement, they were not supposed
to make any money! HP didn't back the projects with experience and planning,
hence the 'hobby' label I gave it - that will lead to failure of any
project.
Add more yourself.
Regards
Steen
Bhuvanesh
> "Steen Schmidt" <ssch...@nospam.dk> wrote in message news:<HdVh8.222$Z25....@news000.worldonline.dk>...
> > That 4MHz piece of hardware is ancient, and was (hopefully) merely a
> > stepping stone for a better calculator. The HP49G is not much more than
> > HP48GX+www.hpcalc.org.
>
> Never considered it this way, but now that you said it...
I know! TI and HP power users are getting closer together this way ;-)
> > Step-by-step? Non-existent. TI-like SOLVER. Not by a longshot.
>
> Well Steen, Step-by-Step is "somehow" existent, but it this "somehow"
> is not understandable by most of the people for using it. I think the
> makers assumed that *everyone* that uses the machine will also have
> the computer/algorithm knowledge that is necessary to interpret the
> starnge messages of the Step-by-Step feature.
Something like the "trace" messages displayed by computer CAS's,
showing what the algorithm is doing? Those can't possibly be used by
newbies. While I am interested in computer algebra, I cannot expect
everyone to be, and sometimes, when you are using the CAS as a tool to
achieve something else, it might be best to abstract away from the
algorithms used.
> Yeap, exactly. After a long (and very excited) time of using it, I get
> the feeling that in the HP49G we have a battle of the OS vs. the CAS.
> Those two contradict each other and fight for the right to take
> control. When two elephants are fighting, the victims are the ants
> (we), as people in India say. (Bhuvanesh ? ;-) )
;-) Well, at least some of the bugs were fixed, which HP by itself
would probably not have done.
> > No piece of software is better than its design specification.
Hehe, I see that here at work every day :-)
> > When software
> > is not documented, you can't fix bugs, and noone else can work on it after
> > you quit or get fired.
That too, and documentation also serves as some form of assurance.
When you use documented features, you can be assured that you will be
able to continue using them in a future version (although they could
be deprecated).
> > If software is not coded in highly specific modules,
> > you can't write a good test specification, and hence can't test it properly.
> > If you can't, won't or don't have the time for testing, the software is a
> > failure, since it'll most definetely contain bugs that'll take a long time
> > to find and be near impossible to fix. When old bugs are left in the code,
> > while newer ones are fixed, you risk introducing secondary bugs again. This
> > cycle will kill your project.
Also, while the developer(s) can provide good test cases (because
he/they wrote the code), someone else needs to test the software too.
> Now Steen, you mean they (HP) torpedoed the project themselves right
> from the start? Why should they do that?
No idea... corporate people do weird things sometimes.
Regards,
Bhuvanesh.
Jean-Yves Avenard
I'm surprised with your past few messages.
I've been used to read fair, well balanced and objective comments. Now it
just sounds like an eager teenager.
Jean-Yves
"Steen Schmidt" <ssch...@nospam.dk> wrote in message
news:aG2i8.4353$Bo.4...@news010.worldonline.dk...
> I don't know why, but they did. The ACO didn't get any funding (economical
> as well as knowledgewise). Many of the promising projects were abandoned
> without obvious reason. The ACO was closed, presumably because they didn't
> make any money - it was a *development* departement, they were not
supposed
> to make any money! HP didn't back the projects with experience and
planning,
> hence the 'hobby' label I gave it - that will lead to failure of any
> project.
That is total crap.
ACO was not a R&D only department, and of course they had fundings ...
>
> Add more yourself.
>
Jean-Yves
Steen Schmidt
> I'm surprised with your past few messages.
Then argue.
> I've been used to read fair, well balanced and objective comments. Now it
> just sounds like an eager teenager.
You disappoint me, Jean-Yves. Can't take the heat?
> That is total crap.
Your choice of words is really astonishing. Will you at least pretend to be
nice - this is a public forum.
> ACO was not a R&D only department, and of course they had fundings ...
Correct me if I'm wrong, but that is what have always been said. If that is
not true, you have all complained about nothing.
Did you or did you not have to sell the 48G+ & HP6S to raise money for the
HP49G development? You have said so, so HP didn't give you any money to play
with, hence they didn't believe in the project. How much Corvallis-knowledge
have you been introduced to? Nothing, have I been told by Cyrille, so I hope
that's the truth. An effect of this can be witnessed if you tell how you got
hold of the old HP calc ROMs. That wasn't exactly through HP, was it?
Now, please respond to my statements, instead of just throwing mindless
accusations at me.
Regards
Steen
Steen Schmidt
> I'm a little bit surprised to read this message from you because
> it is not objective anymore.
I think it is objective. I'll provide my arguments below.
> Sure we had not enough time to
> release a stable ROM at the beginning (but the current 1.19-6 release
> is fairly stable),
ROM v1.19.6 isn't a release ROM, it's a beta. Generally speaking, no user
should use anything but ROM v1.18. ROM v1.18 is a disaster, stability wise,
and practically not usable. Where is ROM v1.20?
If v1.19.6 was relabeled v1.20, and made a release ROM I'd be happy. As of
now, the HP49G does not have a stable release ROM.
> a faster processor would have been greatly
> appreciated
You don't find the processor a roadblock? Do you find the hardware
satisfactory as a final calculator from HP?
> as well as inclusion of the documentation available
> on the web page of Renee De Graeve.
The documentation was a must, and the lack of it probably the major reason
that the HP49G sold so badly. You know how many have complained about the
docs. How could, for example, a former TI-using
student ever figure out to use the '49?
Of course it's not a problem for you and me, but for 99,99% of the customers
it is.
> The HP49 is however a nice
> calc, and we have introduced some concepts that are now imitated
> by TI or TI fans like step-by-step (which exists despite your comment)
Is the step-by-step usable you think? I think not. It's a sales gimmick in
my opinion. It's not even described anywere when S-B-S works and when it
doesn't. I have found these commands to work with
S-B-S on the '49 (S-B-S is on, VX=X:):
The notation is;
- 'argument(s)' COMMAND -> "displayed s-b-s response, next line, another
line..." 'result' (comment(s))
Integration (INTVX, RISCH, FOURIER & DESOLVE sometimes)
Examples:
- 'SIN(X)' INTVX -> '-COS(X)' (No explanation shown)
- 'SIN(U)' INTVX -> 'SIN(U)*X' (No explanation shown)
- 'SIN(U)*X' INTVX -> "SIN(U)*X, Rational fraction, X" 'SIN(U)*(1/2*X^2)' (X
was a rational fraction - is the student familiar with the term 'rational
fraction', and how this applies to 'X'? There
is no further explanation on how the '(1/2*X^2)' part was derived)
- 'EXP(X-2)*5*X' INTVX -> "EXP(X-2)*5*X, Linearizing, 5*X*EXP(X-2)"
'(5*X-5)*EXP(X-2)' (No help. How was it solved? Is the student familiar with
linearization?)
- 'EXP(1/(X-2))*LN(X-2)' INTVX -> "EXP(1/(X-2))*LN(X-2), Int [u'*F(u)] with
u=X-2" "Risch alg. of tower, {'LN(X)' 'EXP(1/X)' 'X'}" "No closed form,
LN(X)*EXP(1/X)" 'INT(LN(Xt)*EXP(1/Xt):Xt:X-2)'
(At least the substitution is shown, but you keep using 'X' instead of 'U'
or 'u'. Is the student familiar with the Risch algorithm? What is alg. short
for? What is a tower, and has it got anything
to do with that odd list shown?)
- 'SQRT(X^2-1)' INTVX -> "SQRT(X^2-1), Square root, SQRT(X^2-1)" "Rational
fraction, 0" "Rational fraction, (-1/2)*(2/-(2*X))"
'X/2*SQRT(X^2-1)+1/2*LN(SQRT(X^2-1)-X)' (Square root was surely
pointed out. Rational fraction, 0? What does that mean? And where did
'(-1/2)*(2/-(2*X))' come from? It seems like a mystery how the calc got to
the answer)
- 'SIN(X)/X' INTVX -> "SIN(X)/X, Risch alg. of tower, {'SIN(X)' 'X'}"
"Invalid unary operator, SIN(X)/X" 'INT(SIN(Xt)/Xt:Xt:X)' (What's with the
error? Is the student familiar with the term 'unary
operator'?)
- 'X^X' INTVX -> "EXP(X*LN(X)), Risch alg. of tower, {'EXP(X*LN(X))' 'X'}"
"EXP>EXP/LN not handled, EXP(X*LN(X))" 'INT(EXP(Xt*LN(Xt)):Xt:X)' (I never
found out what was not handled...)
Matrix operations (INV, RANK, REF, rref, RREF, IMAGE, KER, PMINI, BASIS &
LINSOLVE):
Examples:
- [[-5 -2][2 0]] INV -> "L2=5*L2--2*L1, [[-5 -2 1 0][2 0 0 1]]" "L1=2*L1-L2,
[[-5 -2 1 0][0 -4 2 5]]" "Reduction result, [[-10 0 0 -5][0 -4 2 5]]" [[0
'1/2']['-1/2' '-5/4']] (Is the student
familiar with the 2x4 matrices suddenly used in the calculations? What is
the Ln representation?)
- [[4 5][-7 0]] RANK -> "L2=4*L2--7*L1, [[4 5][-7 0]]" "Reduction result,
[[4 5][0 35]]" 2 (Again, where's the information about rank calculation?)
- [[4 5][-7 0]] REF -> "L2=4*L2--7*L1, [[4 5][-7 0]]" "Reduction result, [[4
5][0 35]]" [[1 '5/4'][0 1]] (How can this be the same as in the RANK case?)
- [[4 5][-7 0]] rref -> "L2=4*L2--7*L1, [[4 5][-7 0]]" "L1=7*L1-L2, [[4 5][0
35]]" "Reduction result, [[28 0][0 35]]" Pivots:{7 1. 4 1.} [[28 0][0 35]]
(A little change, but it's still not very
obvious from this how to calculate the row reduced echelon form)
- [[4 5][-7 0]] RREF -> "L2=4*L2--7*L1, [[4 5][-7 0]]" "L1=7*L1-L2, [[4 5][0
35]]" "Reduction result, [[28 0][0 35]]" [[1 0][0 1]] (Not much different
from rref)
- [[4 5][-7 0]] IMAGE -> "L2=4*L2-5*L1, [[4 -7][5 0]]" "L1=5*L1--1*L2,
[[4 -7][0 35]]" "Reduction result, [[20 0][0 35]]" {[1 0] [0 1]} (I'm
starting to see the picture. Is this much experimenting
necessary? Is it expected by an average student?)
- [[4 5][-7 0]] KER -> "L2=4*L2--7*L1, [[4 5][-7 0]]" "L1=7*L1-L2, [[4 5][0
35]]" "Reduction result, [[28 0][0 35]]" {}
- [[4 5][-7 0]] PMINI -> "L2=L2-4*L1, [[1 0 0 1 1][4 5 -7 0 X][-19
20 -28 -35 X^2]]" "L3=L3--19*L1, [[1 0 0 1 1][0 5 -7 -4 X-4][-19 20 -28 -35
X^2]]" "L3=L3-4*L2, [[1 0 0 1 1][0 5 -7 -4 X-4][0 20
-28 -16 X^2+19]]" "Reduction result, [[1 0 0 1 1][0 5 -7 -4 X-4][0 0 0 0
X^2-4*X+35]]" [[1 0 0 1 1][0 5 -7 -4 X-4][0 0 0 0 X^2-4*X+35]] (I can't
really link the different steps togteher)
- [[4 5][-7 0]] BASIS -> "L2=4*L2--7*L1, [[4 5][-7 0]]" "L1=7*L1-L2, [[4
5][0 35]]" "Reduction result, [[28 0][0 35]]" {[1 0] [0 1]} (Same steps as
for RREF, but different result. How can that be?)
- [4*X+5*Y=-6 -7*X=0] [X Y] BASIS -> "L2=4*L2--7*L1, [[4 5 6][-7 0 0]]"
"L1=7*L1-L2, [[4 5 6][0 35 42]]" "Reduction result, [[28 0 0][0 35 42]]"
{[4*X+5*Y=-6 -7*X=0] [X Y]} Specific:{28 35 7 1. 4 1.} [X=0 Y=-6/5]
Integer operations (IABCUV, ICHINREM, IDIV2, IEGCD, IQUOT & IREMAINDER):
Examples:
- 8 5 6 IABCUV -> "[z,u,v]: z= u*8 + v*5, [8 1 0], [ 5 0 1 ] *-1" "[ 3
1 -1 ] *-1" "[ 2 -1 2 ] *-1" "[ 1 2 -3 ] *-2" "Result:[ 1 2 -3]" 12 -18
(What is the connection between the result vector and the real result? I
can't see the meaning of the intermediate steps either)
- [4 -6] [7 1] ICHINREM -> "[z,u,v]: z= u*-6 + v*1, [-6 1 0], [ 1 0 1 ]
*--6" "Result:[ 1 0 1 ]" [4 -6] (Again, I wouldn't exactly call this
obvious)
- 25 63 IDIV2 -> "25|63, |__________, 25|0" 0 25 (This time, s-b-s is merely
a division line on the display, that much I gather. That's not much help, is
it?)
- 85 12 IEGCD -> "[z,u,v]: z= u*85 + v*12, [85 1 0], [ 12 0 1 ] *-6" "[ 1
1 -7 ] *-12" "Result:[ 1 1 -7 ]" 1 1 -7 (This time there is a relationship
between the result vector and the real result. The rest is foggy though)
- 10 3 IQUOT -> "10|3, |__________, 1|3" 3 (Same information as in the IDIV2
case)
- 10 3 IREMAINDER -> "10|3, |__________, 1|3" 1 (Again)
Polynomial operations (DIV2 & PARTFRAC):
Examples:
- 'X^3-5*X+6' '5*X^2-15' DIV2 -> "Division A=BQ+R, A: { 1 0 -5 6 }, B: { 5
0 -15 }, Q: { '1/5' }, R: { 0 '-5--15*(1/5)', Press a key to go on" "Q: {
'1/5' 0 }, R: { '-5--15*(1/5)' 6, Press a key to go on" Q:'1/5*X'
R:'(-5--15*(1/5))*X+6' (This is much better, although there is more
confusion with the keys/softmenu. The layout of Q & R could also be better)
- '(X^2-4*X-3)/8' PARTFRAC -> '(X^2-4*X-3)/2^3' (This is not split into
partial fractions in s-b-s mode. When s-b-s is disabled, the '49 return
'1/8*X^2+-1/2*X+-3/8')
- '(X+12)/2' PARTFRAC -> '(X+12)/2' (This is, as many other expressions,
also not split into partial fractions in s-b-s mode. When s-b-s is disabled,
the '49 return '1/2*X+6')
- '(4*X-12)/(X^2+6*X)' PARTFRAC -> "Factorizing:(4*X-12)/(X^2+6*X),
Deno.:(X*(X+6)), Bezout AU+BV=:(4*X-12), A:(X+6), B:X"
"((2*X-6)/3*(X+6)-(2*X-6)/3*X)/(X*(X+6))" '-2/X+6/(X+6)' (Pretty messy
infromation.)
- '(281*X+1086)/168' PARTFRAC -> "Factorizing:(281*X+1086)/168,
Deno.:(2^3*3*7), Bezout AU+BV=:(281*X+1086), A:7, B:24"
"(281*X+1086)/24*(2^3*3)/(2^3*3*7)" '(281*X+1086)/24*(2^3*3)/(2^3*3*7)'
(Again, not split into partial fractions. When s-b-s is disabled, the '49
return '281/168*X+181/28')
In my opinion, PARTFRAC is useless with s-b-s mode enabled. I have written
this to you since the first ROMs, and I know you disagree. I'd be glad to
know what everybody else thinks?
I'd also be glad to know if everyone here understands the outpur when s-b-s
is on. Is it usable?
> or EQW.
The performance and novelty of the EQW is superb. That's also pretty much
the only feature I can find, that is special and working flawlessly. That,
and maybe MASD (almost flawless). The GUI on the '49 is also pretty fast.
> And the fact that people continue to ask whether they should buy
> a TI or an HP is a sign that it is in the competition.
Yes, because the people asking trust the papers. When they try the real
thing, they are rarely satisfied. Haven't you noticed that?
> I believe
> that handhelds with e.g. linux OS have a much better future (and
> that's not against a calculator with a keyboard: there are still
> keyboards for PDAs).
I still need to see the advantage of current Linux implementations. As I see
it, Linux has its force in simplicity. Many implementations today are still
bloated. Where I work, we use Linux for many
purposes, mainly because it's free. Where we have mission critical systems,
we do not use Linux (or Windows for that matter). For CPUs that are
interesting in current handheld devices, there still
exist many specialized OS'es. They are expensive, but very stable kernels.
C/C++/EC++ compilers also exist for almost every CPU nowadays, so there is
not much need for Java.
Regards
Steen
Jean-Yves Avenard
"Steen Schmidt" <ssch...@nospam.dk> wrote in message
news:3X9i8.796$Z25....@news000.worldonline.dk...
>
> You disappoint me, Jean-Yves. Can't take the heat?
Of course I can, I even bought a pair of sun glasses
>
> > That is total crap.
>
> Your choice of words is really astonishing. Will you at least pretend to
be
> nice - this is a public forum.
>
But it is.
You make incorrect statement regarding how HP was funding its department.
ACO had a budget, and quite a big one. Otherwise how can you pay so many
engineers ?
I can't release any detailed information otherwise I will see HP lawyers at
my door again.
The HP49 has been correctly funded and so did Xpander. Xpander (and any new
development on the HP49) has been cancelled because HP focused on a
different global strategy which didn't involve education anymore. you can
always argue this decision but that's a different story.
> Correct me if I'm wrong, but that is what have always been said. If that
is
> not true, you have all complained about nothing.
No. What has been said is that after the official release of the HP49, all
the engineers working on the HP49 have been assigned on different projects
and that's perfectly normal. Not that I liked it, it just what usually
happen. So all the development made after ROM 1.18 has been made thanks to
the free time of many engineers.
>
> Did you or did you not have to sell the 48G+ & HP6S to raise money for the
> HP49G development? You have said so, so HP didn't give you any money to
play
That is incorrect.
The HP6S project has been started after the HP49G. It hasn't been a big
success either. I don't think you can buy a 6S anymore. Last time I've
checked it was not in the HP catalog.
> with, hence they didn't believe in the project. How much
Corvallis-knowledge
> have you been introduced to? Nothing, have I been told by Cyrille, so I
hope
The only reason is that nobody from the old Corvallis team was working for
HP at that time except Jim Donnelly, and Jim's involvment in the HP48 was
minimum as he worked on the HP library card (periodic table, solver etc..).
I've met Diana Byrne on this matter just 2 weeks before she left for TI.
We had access to all the ressources available at the time.
> that's the truth. An effect of this can be witnessed if you tell how you
got
> hold of the old HP calc ROMs. That wasn't exactly through HP, was it?
The ROM was incomplete that's correct, but the main problem was the
incomplete HP Tools which was not available on the new HP unix platform. But
what can you expect from a 10 year old project that has been relocated 2
times over the world.
>
> Now, please respond to my statements, instead of just throwing mindless
> accusations at me.
>
I am not.
I still believe that the HP49 was developped during the good time of HP. It
started to deteriorate AFTER (about 2 years ago)
Jean-Yves
Parisse Bernard
>
> ROM v1.19.6 isn't a release ROM, it's a beta. Generally speaking, no user
> should use anything but ROM v1.18. ROM v1.18 is a disaster, stability wise,
> and practically not usable. Where is ROM v1.20?
>
> If v1.19.6 was relabeled v1.20, and made a release ROM I'd be happy. As of
> now, the HP49G does not have a stable release ROM.
>
I don't see any difference because in both case the user has to upgrade
his calc by himself, and if he want to upgrade then he probably knows
that the beta 1.19.6 is much more stable than the so-called stable 1.18.
> The documentation was a must, and the lack of it probably the major reason
> that the HP49G sold so badly. You know how many have complained about the
> docs. How could, for example, a former TI-using
>
> student ever figure out to use the '49?
>
> Of course it's not a problem for you and me, but for 99,99% of the customers
> it is.
>
I agree about the doc, but it seems that unfortunately people do not
read the docs hence managers decided to spare money providing doc
on the web (and if you look at TI it's the same as HP).
> Is the step-by-step usable you think? I think not. It's a sales gimmick in
> my opinion. It's not even described anywere when S-B-S works and when it
> doesn't.
step by step is something that depends on the country habits to learn
maths. It is well adapted to the French way, I don't know other
countries
way.
>
> Integration (INTVX, RISCH, FOURIER & DESOLVE sometimes)
> Examples:
>
> - 'SIN(X)' INTVX -> '-COS(X)' (No explanation shown)
And what whould you expect?
> - 'SIN(U)' INTVX -> 'SIN(U)*X' (No explanation shown)
same!
> - 'SIN(U)*X' INTVX -> "SIN(U)*X, Rational fraction, X" 'SIN(U)*(1/2*X^2)' (X
> was a rational fraction - is the student familiar with the term 'rational
> fraction', and how this applies to 'X'? There
>
> is no further explanation on how the '(1/2*X^2)' part was derived)
Yes a student should know the main methods for integration before using
the s-b-s feature. The calc shows what method it uses.
> - 'EXP(X-2)*5*X' INTVX -> "EXP(X-2)*5*X, Linearizing, 5*X*EXP(X-2)"
> '(5*X-5)*EXP(X-2)' (No help. How was it solved? Is the student familiar with
> linearization?)
yes, he should. The calc does not replace the teacher.
Same applies on the other examples.
>> I believe
>> that handhelds with e.g. linux OS have a much better future (and
>> that's not against a calculator with a keyboard: there are still
>> keyboards for PDAs).
>
>
> I still need to see the advantage of current Linux implementations. As I see
> it, Linux has its force in simplicity. Many implementations today are still
> bloated. Where I work, we use Linux for many
>
> purposes, mainly because it's free. Where we have mission critical systems,
> we do not use Linux (or Windows for that matter). For CPUs that are
> interesting in current handheld devices, there still
>
> exist many specialized OS'es. They are expensive, but very stable kernels.
> C/C++/EC++ compilers also exist for almost every CPU nowadays, so there is
> not much need for Java.
>
I don't think specialized OS will survive. Probably only pocket PC,
linux, even
palmOS is not sure. We'll see. One of the main advantage of e.g. linux
is that
you develop your app for the desktop and just cross-compile it for the
PDA
it's a 2 minutes work to port. BTW I don't use java, linux is not a
synonym
to java by far.
dan watts
recommended the RPN way, and I'm so glad I followed his advice. It
enabled me to score significantly higher marks than my classmates. In
studying precalculus I find the Equation Writer and the Step-by-Step
mode very useful. I love the solid feel of the 49g keyboard. I printed
and bound Renee De Graeve's "Calcul formel et Mathematiques avec la
HP49G en mode algebrique", translated into English by Ivan Bertolotti,
from the website www-fourier.ujf-grenoble.fr/~degraeve/usflan.pdf, and
have found it a terrific resource. Nick Karagiaouroglou's Trig
Marathon is a tremendous resource, as will be the upcoming Complex
Number Marathom and Calculus Olympics. For me, HP is the place to be.
I took a 4-hour electronics exam recently with a public utility. Their
wording was "a calculator will be the only study aid allowed during
the test." I hadn't had any exposure to digital circuits, so I had 5
weeks to learn enough digital to pass the exam. I downloaded Robin
Getz's Digitician from hpcalc.org website, which takes Karnaugh map
data in decimal string format and uses the Quine-McCluskey algorthm to
convert to a simplified Boolean expression. This proved to be very
useful. I was able to guide my shaky Boolean abilities and check my
answers with it.
My digital textbook, "Digital Systems" by Ronald Tocci says, "The
following discussion (of the Karnaugh map method) will be limited to
problems with up to four inputs, since even five- and six-input
problems are too involved and are best done by a computer
program."(pg. 122) It is interesting to note that Getz's Digitian,
even while not written in "machine code" and "sacrificing program
memory and speed for low run time memory", is able to solve a 6-input
circuit with 15 highs and 10 don't cares in 8 minutes. On the 3- and
4-input circuits on the test I took, the solve time seemed plenty
fast, maybe around a minute. My wife says I love my calc more than I
love her, and sometimes I'm afraid she's right!
I have a question for Steen. Do you really think you are going to able
to bring a superior calc to market any time soon at under $400? Are
you going to be able to get it on the shelf at the Best Buys and
Circuit Cities and Office Maxes and their international equivalents
next to the TIs? Are you going to be able to break TI's stranglehold
on the schools? I sincerely hope you are successful in doing so. But
it seems to me the odds of your success are maybe one in a thousand.
And if you are successful, are you going to give your HP predecessors
due credit?
Steen Schmidt
> his calc by himself,
You don't see a difference between a beta and a release ROM? Are you happy,
if your car needed a beta ECU program, installed at the first service, to
work? Would that car manufacturer be as professional as the one that had
stable ECU ROMs from the beginning, in your opinion? Does HP endorse the use
of v1.19.6? No. A user should be very confident in himself, if he was to
load a beta ROM into his calc. You can't get any support, and the warranty
is viod. If the calc dies on your, with a beta ROM loaded, you're on your
own. How many, shelling out USD 200 for a calc, are willing to run that
risk? Not many outside this group are, and even some here aren't.
You can't just tell someone to go load ROM v1.19.6 - it's bad business and
unprofessional. As far as I'm concerned, HP does not back any stable ROMs
for the '49. Re-label ROM v1.19.6 to v1.20, and you have yurself a very good
release ROM.
> and if he want to upgrade then he probably knows
> that the beta 1.19.6 is much more stable than the so-called stable 1.18.
No, he probably doesn't. The manual says that there may be more ROMs, and
that you can upgrade the calc. 99% of the customers does not know which
effect a ROM upgrade can have, hence they don't know that a ROM upgrade can
maybe solve their stability problems. I know many that complain about the
'49 crashing, and most use v1.10 or v1.18. The calc should at least function
properly with the OEM loaded ROM. The '49 does not.
If we take a brand new Plextor CDRW for a PC as an example. You come home,
and you find out that it cannot write on a certain brand of blank CDs that
you've bought. You go read the Plextor FAQ (because yu're proficcient with
the internet, as you may not even be, if you buy a HP49G), and it says to
upgrade your ROM (such a Plextor burner also has Flash technology). You then
load that new ROM, and your media gets burned fine now. You then discover
that the new ROM make the eject button not work. The Plextor FAQ does not
even list this problem, nor has it a solution for it. Some guys that worked
for Plextor a while back, have released a couple of beta ROMs for that
Plextor burner. One of the latest betas fix the problem with the eject
button, while keeping the new media working. You don't know that this ROM
exist, and Plextor does not metion it on their site. You have, maybe, seen
pages found with Google that mention some beta ROMs for your CDRW, but they
are not endorsed nor supported by Plextor, so you don't dare use them. I'd
take that burner back!
> I agree about the doc, but it seems that unfortunately people do not
> read the docs hence managers decided to spare money providing doc
> on the web (and if you look at TI it's the same as HP).
Docs on the web is fine, but that documentation wasn't there at release, was
it? How long was it, before we could point people towards it?
> step by step is something that depends on the country habits to learn
> maths. It is well adapted to the French way, I don't know other
> countries way.
There we have many of the porblems about the HP49G. It's a French
calculator. I thought mathematic was universal? Are French students used to
find " "SQRT(X^2-1), Square root, SQRT(X^2-1)" "Rational fraction, 0"
"Rational fraction, (-1/2)*(2/-(2*X))"
'X/2*SQRT(X^2-1)+1/2*LN(SQRT(X^2-1)-X)'" in their math books? I think not.
You could've done an effort to make the s-b-s feature more universal. What
about more explanations? Why are some commands called for example DEDICACE
and DROITE? You know that only French people will understand what they are
for. If the user wants s-b-s help, performance is not a problem. Why not
really show the steps?
> > - 'SIN(X)' INTVX -> '-COS(X)' (No explanation shown)
> And what whould you expect?
For example: "S(SIN(X),X) -> '-COS(X)', by elementary function table
look-up."
Where 'S' is the integration "character".
> > - 'SIN(U)' INTVX -> 'SIN(U)*X' (No explanation shown)
> same!
For example: "S(F(U),X) -> 'F(U)*X', Any term not involving the integration
variable is considered a constant term U, and hence merely multiplied by the
integration constant."
These different output messages in the s-b-s mode should then be described
in the manual, with examples so the user can understand what's happening.
When I first bought my HP48 series calculator, I learned more math than I
knew before from the manual. Many of the things it can I wasn't been tought
yet in school. Ididn't know how to get by improper integration for example.
I learned how to by the HP48 manual.
> > - 'SIN(U)*X' INTVX -> "SIN(U)*X, Rational fraction, X"
'SIN(U)*(1/2*X^2)' (X
> > was a rational fraction - is the student familiar with the term
'rational
> > fraction', and how this applies to 'X'? There is no further explanation
on how the
> > '(1/2*X^2)' part was derived)
> Yes a student should know the main methods for integration before using
> the s-b-s feature. The calc shows what method it uses.
Where is it shown in the manual, which methods the calc can use, and how
they work? One could use more display real estate, and more words to point
out which parts of the expression was involved at what time in the solution.
One could seperate all terms, and take them one at a time. Then
factors/constants could be extracted and handled like this:
"S(F(U)*F(X)) -> F(U)*S(F(X)), Constants are merely factored out of the
integration and multilplied onto it afterwards. Any factor not involving the
integration variable is considered a constant factor F(U)"
This should of course also be explained in the manual, with examples.
THEN you could do something like "'X' is a polynomial expression, and
integrated term by term like this: S(a*X^n) -> a/(n+1)*X^(n+1)"
Had the above been a true fraction, like '(X^2+X-6)/(X^2+2*X-24)', do you
expect all students to be able to integrate that? You could have shown this
then:
"'(X^2+X-6)/(X^2+2*X-24)' is a rational fraction, and a general rational
fraction is integrated one term at a time, of its partial fraction
decomposition..........something"
And then shown step-wise, first the PARTFRAC result, then 'each term' to
'a/b*LN(X+c)' and then....
> > - 'EXP(X-2)*5*X' INTVX -> "EXP(X-2)*5*X, Linearizing, 5*X*EXP(X-2)"
> > '(5*X-5)*EXP(X-2)' (No help. How was it solved? Is the student familiar
with
> > linearization?)
> yes, he should. The calc does not replace the teacher.
Of course it does not, but that is not the question. What if you don't have
a teacher nearby - shouldn't the manual be helpful then? There is more than
linearization to the solution of the above example too - you could show how
the constant (5) was handled, and that you used the property
S(x*EXP(x+n)) -> (X-1)*EXP(X+n) (or whichever you used).
What about my other examples, are they obvious you think?
> I don't think specialized OS will survive. Probably only pocket PC,
> linux, even palmOS is not sure.
What about WinCE? Also - as long as Intel makes StrongARM processors, most
ARM kernels will survive too.
> We'll see. One of the main advantage of e.g. linux
> is that you develop your app for the desktop and just cross-compile it for
the
> PDA it's a 2 minutes work to port.
It'd be the same with anything written in C/C++/EC++. Not that many parts of
the kernel needs to be in assembly, with todays embedded processors.
> BTW I don't use java, linux is not a
> synonym to java by far.
Did I say that? Linux is an OS kernel, and Java is a programming language. I
compared C++ with Java, not Linux with Java.
Regards
Steen
Steen Schmidt
> ACO had a budget, and quite a big one. Otherwise how can you pay so many
> engineers ?
Of course the salaries were in the budget, but I'm also talking about
technological aid, not merely echonomical.
> I can't release any detailed information otherwise I will see HP lawyers
at
> my door again.
Again? :-)
> The HP49 has been correctly funded and so did Xpander. Xpander (and any
new
> development on the HP49) has been cancelled because HP focused on a
> different global strategy which didn't involve education anymore. you can
> always argue this decision but that's a different story.
Ok, but that's a first for me. I guess nobody here have ever heard this
version before?
> No. What has been said is that after the official release of the HP49, all
> the engineers working on the HP49 have been assigned on different projects
> and that's perfectly normal. Not that I liked it, it just what usually
> happen. So all the development made after ROM 1.18 has been made thanks to
> the free time of many engineers.
So, HP abandoned the HP49G, before a stable ROM could be released? That's
not backing the ACO - that was all there was in my first statement. I said
that HP didn't ever back the ACO, and I'm right. If they did, the products
would be finished, and in good working order.
> That is incorrect.
> The HP6S project has been started after the HP49G.
But it was released after. What about the HP48G+? YOU have always stated
that this was merely to raise money - why, if HP paid everything?
> We had access to all the ressources available at the time.
I guess no one at HP has the ROM sources for any HP calculator then? That
seems very odd. Whatever the cause, when I talked with Cyrille about this,
HP goodwill was not a major concern. It was practically non-existent. If it
was him exagerating, then let that be. This was a couple of months before
the "release" of the JX25 - you now where.
> The ROM was incomplete that's correct, but the main problem was the
> incomplete HP Tools which was not available on the new HP unix platform.
Not much was available on that box, right? Not even a LAN connection I was
told.
> But
> what can you expect from a 10 year old project that has been relocated 2
> times over the world.
Sure, but I would expect HP to be able to recreate the project. That's the
responsebility of any project manager, to save enough information to be able
to do just that. That we have done everywhere I have evr worked here in
Denmark - I can only hope it is like that in the rest of the world. Does HP
save precious information now?
> I still believe that the HP49 was developped during the good time of HP.
It
> started to deteriorate AFTER (about 2 years ago)
I thought you didn't agree with me, that HP is gone as we know it? You do, I
see.
HP did start the ACO for a reason, so of course they supported it at first.
HP didn't seem to care though, since no real investment was made - I'm
talking about technological and spiritual backing here, not a few USD.
But you know best, of course. If you say that HP stands 100% completely
behind the ACO in everything, then I won't try to convince you otherwise.
Regards
Steen
Steen Schmidt
> to bring a superior calc to market any time soon at under $400?
It won't be soon, it won't be a calculator, but it will be under USD 400.
> Are
> you going to be able to get it on the shelf at the Best Buys and
> Circuit Cities and Office Maxes and their international equivalents
> next to the TIs?
I don't think so - it wouldn't compete with a Voyager 200, so they'd
probably need to create some new shelfspace, right next to their other
high-tech equipment. Do they have such shelf space currently? I think not,
but it's not important..
> Are you going to be able to break TI's stranglehold
> on the schools?
I hope not, as the current projects aren't exactly for schools. I'm going
back to the customers for old HP handheld products - these are mainly
engineers and geeks. Data/Math/Phys/Chem Students will of course culd use
one, but it'll depend on price if it's attarctive. It'll of course replace
their PDA and much more...
It'll can of course be a very powerful calculator, and only software will
set the limits of what it'll be able to do as a calculator. the display (if
you need one) will be of much better quality than what you expect today. Of
course, the technological step we are at in two years time will tell us what
display to use, but currently we're using a high quality Seiko display. This
will easily fit the bill, so it probably won't get any worse, quality wise.
I can't tell you much of the concept, but you can look at it like a device
with computing power to feed at least one and probably two or more firewire
ports, or to invert a 100x100 matrix in less than one second. It will have
power enough to act as a 8-way (or more) LAN switch if needed. What you need
of I/O capabilities, display, battery power, computing power and so on is
mainly defined by the user. So is the implementation via software (is it to
be a caluclator or an oscilloscope, or maybe both? - the user decides).
Again - expect nothing and you'll be positively surprised.
> I sincerely hope you are successful in doing so. But
> it seems to me the odds of your success are maybe one in a thousand.
> And if you are successful, are you going to give your HP predecessors
> due credit?
The ACO? If anything, it was their demise that started it :-/
Regards
Steen
Nick Karagiaouroglou
> > > It's a failure.
> > Oops! I wouldn't go that far to say that it is a complete failure
> > Steen.
>
> A failed product (couldn't sell on the proposed market), not that much in
> the sense of a failed calculator.
Well, economically perhaps it was, but technologically it has many
(hidden) strengths.
> > > There was no support, and hence HP
> > > didn't want this to succeed.
> > Now Steen, you mean they (HP) torpedoed the project themselves right
> > from the start? Why should they do that?
>
> I don't know why, but they did. The ACO didn't get any funding (economical
> as well as knowledgewise)
I'm also convinced since a long time that the ACO was treated pretty
much like a step daughter by HP. But JYA says "No". Well, perhaps he
has a different opinion about what "treated like a step daughter"
means. I think HP didn't realize what potential is in his mind and in
the mind of BP.
> Many of the promising projects were abandoned
> without obvious reason. The ACO was closed, presumably because they didn't
> make any money - it was a *development* departement, they were not supposed
> to make any money! HP didn't back the projects with experience and planning,
> hence the 'hobby' label I gave it - that will lead to failure of any
> project.
Exactly my thoughts.
> Add more yourself.
What could I add more? When such brilliant people like the ACO team
are treated like that?
Greetings,
Nick.
Nick Karagiaouroglou
> n...@imos-consulting.com (Nick Karagiaouroglou) wrote:
>
> > "Steen Schmidt" <ssch...@nospam.dk> wrote in message news:<HdVh8.222$Z25....@news000.worldonline.dk>...
> > > That 4MHz piece of hardware is ancient, and was (hopefully) merely a
> > > stepping stone for a better calculator. The HP49G is not much more than
> > > HP48GX+www.hpcalc.org.
> >
> > Never considered it this way, but now that you said it...
>
> I know! TI and HP power users are getting closer together this way ;-)
Yeah! See the advantages when there are apparently only disadvantages.
Perhaps HP did the its best for TI/HP-Users-Unification releasing the
HP49G. ;-)
> > > Step-by-step? Non-existent. TI-like SOLVER. Not by a longshot.
> >
> > Well Steen, Step-by-Step is "somehow" existent, but it this "somehow"
> > is not understandable by most of the people for using it. I think the
> > makers assumed that *everyone* that uses the machine will also have
> > the computer/algorithm knowledge that is necessary to interpret the
> > starnge messages of the Step-by-Step feature.
>
> Something like the "trace" messages displayed by computer CAS's,
> showing what the algorithm is doing? Those can't possibly be used by
> newbies. While I am interested in computer algebra, I cannot expect
> everyone to be, and sometimes, when you are using the CAS as a tool to
> achieve something else, it might be best to abstract away from the
> algorithms used.
Yes, I meant messages like "Integrating by parts" or similar, which
are more on the abstract side, without mentioning the exact algorithm
that is used. (Steen gave a lot of examples for what such messages
could look like.)
> > Yeap, exactly. After a long (and very excited) time of using it, I get
> > the feeling that in the HP49G we have a battle of the OS vs. the CAS.
> > Those two contradict each other and fight for the right to take
> > control. When two elephants are fighting, the victims are the ants
> > (we), as people in India say. (Bhuvanesh ? ;-) )
>
> ;-) Well, at least some of the bugs were fixed, which HP by itself
> would probably not have done.
So in this cas the victims were the bugs and not the ants. ;-)
> > > No piece of software is better than its design specification.
>
> Hehe, I see that here at work every day :-)
>
> > > When software
> > > is not documented, you can't fix bugs, and noone else can work on it after
> > > you quit or get fired.
>
> That too, and documentation also serves as some form of assurance.
> When you use documented features, you can be assured that you will be
> able to continue using them in a future version (although they could
> be deprecated).
>
> > > If software is not coded in highly specific modules,
> > > you can't write a good test specification, and hence can't test it properly.
> > > If you can't, won't or don't have the time for testing, the software is a
> > > failure, since it'll most definetely contain bugs that'll take a long time
> > > to find and be near impossible to fix. When old bugs are left in the code,
> > > while newer ones are fixed, you risk introducing secondary bugs again. This
> > > cycle will kill your project.
>
> Also, while the developer(s) can provide good test cases (because
> he/they wrote the code), someone else needs to test the software too.
>
> > Now Steen, you mean they (HP) torpedoed the project themselves right
> > from the start? Why should they do that?
>
> No idea... corporate people do weird things sometimes.
Yes, I must fight with them everyday. ;-)
Greetings,
Nick.
Jordi Hidalgo
> I don't think you can buy a 6S anymore. Last time I've checked it was not
> in the HP catalog.
You get one free when buying a batch of HP Paper (White Male Premium or so).
> > How much Corvallis-knowledge have you been introduced to? Nothing [...]
> The only reason is that nobody from the old Corvallis team was working for
> HP at that time except Jim Donnelly, and Jim's involvment in the HP48 was
> minimum as he worked on the HP library card (periodic table, solver etc..).
Did anyone at ACO read his books, not to learn RPL (of course!) but to learn
how to write good documentation? Did anyone ask Jim about why he spent so much
time working on documentation? Did anyone ask him about his job for the HP-71B
and other obsolete calcs? Did anyone ask him what is so exciting about
collecting HP calcs? Did anyone ask him about not providing an attachable
keyboard or about rubber keys? Did anyone ask him why HP calcs had such a
great reputation?
Was Jim's involvement in the HP48 minimum?
ACO's involvement in traditional HP quality has been null. Moreover, harmful.
HP49 = HP48 - HP Quality + Lots of Easter eggs + (MK + CAS, which were '48 apps)
Regards,
Bye.
Jordi Hidalgo
HPCC member #1046
jo...@tv3mail.com
PS: Did anyone ask Jim about the colour blue? :-)
Cyrille de Brébisson
Well, I happen to agree with Micah, the SATURN is a nice CPU, but you have
to admit it's outdated! it's a 20 year old architecture, based on
requirements that are not valid anymore. As for the software, they are nice
tricks in it, and it's a piece of jewelry, but again, of outdated jewelry.
The functionality are nice (for most of them), so I fully agree to the fact
that it should be rewritten in a more modern language on a more modern CPU
(what about an arm platform?)
regards, Cyrille
"George Tsiros" <gts...@yahoo.com> wrote in message
news:df4089c6.02030...@posting.google.com...
George Tsiros
> Well, I happen to agree with Micah, the SATURN is a nice CPU, but you have
> to admit it's outdated! it's a 20 year old architecture, based on
-Find me a computing system that has so low levels of power
consumption. (even a damn pen(cil) can't last for months!!!)
-The fact that something is old, might mean that it is tested, proven
reliable and worthy.
-Me and a friend are trying to find super-low consumption CPUs. Guess
what. The saturn has proven to be so low power, that only one or two
systems are so well designed that can deliver so low power/MHz... one
of which is actually technology in development *NOW*. (has to do with
saving energy at the logic gate level 8-| )
> regards, Cyrille
i like you, man, thanks for your work on the 49...
Steen Schmidt
> consumption. (even a damn pen(cil) can't last for months!!!)
A prototype motherboard I'm testing a bit on currently, runs more than 70
hours on one battery charge. This is including a high resolution display
(higher res than the TI92+), a keyboard and a very powerful CPU (can't tell
you which, but the pack delivers performance roughly 500-3000 times that of
the HP49G). The LiON cells I'm using right now are around 3500 mAh combined.
We're aiming at at least 200 hours continuos on (without backlight), and
that doesn't seem too unreachable.
Newer technology beats the Saturn in most ways as I see it. No wonder - it's
20 years old, as Cyrille points out.
Regards
Steen
John H Meyers
> No wonder - it's 20 years old.
How old is the 'x86' family?
When is it finally going to cease production?
(along with MS-DOS?)
In other words, if it were desired (which it wasn't),
the Saturn could have been "scaled up" a bit,
just as was that Intel stuff ;-)
But today you'd presumably want to rewrite everything
in more general and portable languages, and you could
still slip a Saturn emulator and some old roms (48 and 49)
into some corner, and retain all existing working software.
Actually, what would even be wrong with a piece of silicon
which has both a new processor and a Saturn on board?
Just as HP built "the latest HP financial calc" into
several newer products, someone could, one way or another,
build several existing HP calcs into any new product,
and thus give potential purchasers (if any :)
an easy upgrade path. But you could take more market share
away from TI if you could also build all the TI8x/92 into it ;-)
[r->] [OFF]
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
Timité Hassan
> >
> > > That is total crap.
I don't think so at least not after what you said !
> > Your choice of words is really astonishing. Will you at least pretend to
> be
> > nice - this is a public forum.
> >
> But it is.
> You make incorrect statement regarding how HP was funding its department.
> ACO had a budget, and quite a big one. Otherwise how can you pay so many
> engineers ?
> I can't release any detailed information otherwise I will see HP lawyers at
> my door again.
> The HP49 has been correctly funded and so did Xpander. Xpander (and any new
> development on the HP49) has been cancelled because HP focused on a
> different global strategy which didn't involve education anymore. you can
> always argue this decision but that's a different story.
What the hell does that mean ?
You say that the HP49 has been correctly funded ?
Then why the Hell have you needed to design and to sell the HP48G+ ?
Why have you wasted some time and some money on this calculator when
you could have directly worked on the HP49 ?
Perhaps even you would have had enough time and money to design a
faster C.P.U and a better controller than the Yorke.
Also if you were correctly funded for the HP49G,how come this
calculator had such a poor documentation and marketing ?
Because for me it is either a lack of money or a lack of competent
people.
> > Did you or did you not have to sell the 48G+ & HP6S to raise money for the
> > HP49G development? You have said so, so HP didn't give you any money to
> play
>
> That is incorrect.
> The HP6S project has been started after the HP49G. It hasn't been a big
> success either. I don't think you can buy a 6S anymore. Last time I've
> checked it was not in the HP catalog.
Ok about the HP6S but you haven't answered about the HP48G+.
Why have you(A.C.O) released the HP48G+ if you didn't need money for
further developpements ?
> > with, hence they didn't believe in the project. How much
> Corvallis-knowledge
> > have you been introduced to? Nothing, have I been told by Cyrille, so I
> hope
>
> The only reason is that nobody from the old Corvallis team was working for
> HP at that time except Jim Donnelly, and Jim's involvment in the HP48 was
> minimum as he worked on the HP library card (periodic table, solver etc..).
> I've met Diana Byrne on this matter just 2 weeks before she left for TI.
> We had access to all the ressources available at the time.
>
> > that's the truth. An effect of this can be witnessed if you tell how you
> got
> > hold of the old HP calc ROMs. That wasn't exactly through HP, was it?
>
> The ROM was incomplete that's correct, but the main problem was the
> incomplete HP Tools which was not available on the new HP unix platform. But
> what can you expect from a 10 year old project that has been relocated 2
> times over the world.
>
> >
> > Now, please respond to my statements, instead of just throwing mindless
> > accusations at me.
> >
>
> I am not.
> I still believe that the HP49 was developped during the good time of HP. It
> started to deteriorate AFTER (about 2 years ago)
This is certainly why it has an outdated hardware and the worse
documentation ever released for an advanced calculator all companies
commbined.
> Jean-Yves
Erwann ABALEA
> What the hell does that mean ?
> You say that the HP49 has been correctly funded ?
> Then why the Hell have you needed to design and to sell the HP48G+ ?
What is the purpose of a company? Make money. That's a good enough reason
to make good people work on easy products, that they'll be able to sell
easily. That way, the margin is much higher than what you can get with a
new product like the HP49 was.
So asking for people to make profit is not unreasonable, even if they have
sufficient money to build a new product. I think the HP48G+ was an easy
way of getting some more money, and make the market wait without going
into the Evil Side (tm).
> Why have you wasted some time and some money on this calculator when
> you could have directly worked on the HP49 ?
> Perhaps even you would have had enough time and money to design a
> faster C.P.U and a better controller than the Yorke.
Silly. The time needed to develop an entire new product including a
totally new hardware is much higher than was was permitted.
--
Erwann ABALEA <erw...@abalea.com> - RSA PGP Key ID: 0x2D0EABD5
-----
D'accord, mais si on se met à utiliser des arguments intelligents dans
ce genre de débat, il devient impossible de discuter.
Vous sombrez dans la facilité.
-+- TS in GNU : La dialectique n'est plus ce qu'elle était.
Cyrille de Brébisson
Well, let us put it that way, the saturn delivers roughly 1.6 Mips / What, a
Intel Strong arm 1100 at 133Mhz delivers 200 Mips / What
This means that if the SATURN (or Yorke) was redesigned now, it would be 125
* faster for the same power usage (or 125 times less power hungry at the
same speed)....
The saturn uses roughly 20mA at 5V, this means 100mW, as a comparison, an
Intel PXA250 at 300Mhz uses 300mW (and, as power consumption is factor of
the square of the speed, this means that the same CPU should disipate the
same power than the saturn at 173Mhz!)!
Now, truth is, performance per mips is only an issue in some embedded
systems, so it is difficult to find a low computing power / low power usage
CPU, but have you look at embedded MIPS or embedded Arm7 platform? you will
find that the Saturn is not that good!
Conservatives estimates 3 years ago showed that the Saturn could be
redesigned to run at 40Mhz with a lower power consuption...
Regards,Cyrille
"George Tsiros" <gTs...@yahoo.com> wrote in message
news:ad04cde6.0203...@posting.google.com...
Cyrille de Brébisson
Well, I am normally on the side of the individual and not of the
corporation, but on this one, I have to admit that you are bulying HP a
litle more than it desearves...
Also, please note that I am not releasing any informaiton here or telling
what happen, just proposing alternative thinking
> You say that the HP49 has been correctly funded ?
Well, it made it, and pretty much as we wanted it to be (appart from the
keys, I know!) which means that the money was there...
> Then why the Hell have you needed to design and to sell the HP48G+ ?
As far as any HP48 expert can say, there is no design needed to create a
HP48G+, it's exacly the same product than a HP48G except for the memory
device which is 128K instead of 32K (well, technicaly, it's the same than a
GX minus the port connector, and the bank switcher). As for the selling
part, well, if it looks like you can make a buck or to at minimal cost, why
not do it? it's called extending a product range....
> Why have you wasted some time and some money on this calculator when
> you could have directly worked on the HP49 ?
A couple of possible answers: Because they needed some time to think about
what they were going to do next, because they were sourcing the budget
(these things takes time in a corporate environment), because they did not
have the staff to do so, because they wanted the peoples to learn first how
to deal with calculators by introducing a simpler product like the 48G+,
because they were looking for partners, maybe, the 49 was already in the
pipe and in design in parallel with the HP48G+, maybe because they had other
product to introduce first that made more marketing sence, maybe they were
trying to scope the work for the HP49, working on the schedule
> Perhaps even you would have had enough time and money to design a
> faster C.P.U and a better controller than the Yorke.
Well, ASIC design time is in years from starting the work to first
prototypes. the 49 was introduced barely a year after the ACO was created,
Even with an experienced team, it would not have been possible to redesign
the Yorke in such a small timeframe.
> Also if you were correctly funded for the HP49G,how come this
> calculator had such a poor documentation and marketing ?
> Because for me it is either a lack of money or a lack of competent
> people.
well, it could also be lake of time, lake of people all together, corporate
heaviness, underscoping of tasks, lake of comunication between teams, bad
(or changing) requierements.
Also, remember that the 49 was the first product of a young division and
team, a learning product... and they definitely made beginner mistakes...
> Ok about the HP6S but you haven't answered about the HP48G+.
> Why have you(A.C.O) released the HP48G+ if you didn't need money for
> further developpements ?
maybe because HP wanted to keep a competitive advantage by having calculator
with more memory than competitors? having great product is not all, you also
need to keep watching for competition
> > I am not.
> > I still believe that the HP49 was developped during the good time of HP.
It
> > started to deteriorate AFTER (about 2 years ago)
>
> This is certainly why it has an outdated hardware and the worse
> documentation ever released for an advanced calculator all companies
> commbined.
well, some 4 operation calculator actually DO have worth documentation than
the 49 (amybe because of this Corean to Tawanees to English to French
translation process :-)
Regards, Cyrille
Tom Garcia
"Steen Schmidt" <ssch...@nospam.dk> wrote in message news:<8ici8.1166$Z25....@news000.worldonline.dk>...
> If v1.19.6 was relabeled v1.20, and made a release ROM I'd be happy. As of
> now, the HP49G does not have a stable release ROM.
How can you argue against the quality of a product because of its
name? ("beta", whatever) That's classic judgement of a book by its
cover.
To decide which software to install, on all my systems from servers to
calculators, I observe the actions of expert users and do my own
evaluation. I don't blindly trust a company. Two days after buying my
HP49, I installed 1.19.6. It was completely painless.
Do you install Microsoft's latest OS the day of its gold release,
because Microsoft says it's release quality? Do you apply exactly the
patches it tells you, when it tells you? And if you do, is it because
they will provide support and pay damages if it goes wrong? (Ha!)
While corporate incompetence may have affected technical quality, I'm
not going to pre-judge the technical quality *because of that
incompetence*. As far as I'm concerned, the ideas and overall design
of the HP calculator software are excellent (thank you to all teams
involved, whether in Aus, Asia or USA).
> > a faster processor would have been greatly
> > appreciated
>
> You don't find the processor a roadblock? Do you find the hardware
> satisfactory as a final calculator from HP?
I switched from TI-89 to HP49, and found the former marginally faster
for some operations. Certainly not enough of an improvement to
influence which one I use. I agree that *no* calculator has a state of
the art processor. For those moments when your calculator is a
replacement for Mathematica/Mathcad/whatever, that's relevant. For
now, when I need that power, on goes the laptop.
An ARM-based calculator would be nice instead for on-board power, but
not if I have to recharge at the day's end. I know ARM assembler, but
I'm not sure if I want to learn another :-).
> The documentation was a must, and the lack of it probably the major reason
> that the HP49G sold so badly. You know how many have complained about the
> docs. How could, for example, a former TI-using
> student ever figure out to use the '49?
Well, ok, let me tell you how I did it, since I'm in exactly that
category.
1. I downloaded the user's guide and read it cover to cover. About
half way, I changed to RPN and kept it there since. It's not that hard
to understand a stack machine.
2. I downloaded the AUG and read most of it cover to cover, excluding
the command reference, which I use as a reference :-).
3. I downloaded various other docs from hpcalc.org, e.g. a UserRPL
tutorial, which I worked through during spare moments. I grabbed some
other docs that I will/have read when I'm ready, such as a SysRPL
manual.
4. I read the newsgroup. From it and other documents, I noted some
sexy apps to install, such as Emacs, which made my life easier.
In fact, it's not dissimilar to the way I learn how to use FreeBSD or
Linux. And that's exactly what I wanted.. a few helpful and
enthusiastic people in a community are infinitely better than one huge
company which gives you reams of faceless documentation then leaves
you out in the cold (Microsoft and TI are prime examples of this).
> Of course it's not a problem for you and me, but for 99,99% of the customers
> it is.
I find it hard to grasp that anyone who needs the power of an HP49
can't find adequate information, assuming Internet access. The HP49
isn't a lazy person's calculator, nor a high school calculator (no
matter what HP tried to make it).
How about this on the first page of the manual: "Most documentation is
provided on line. See http://www.hpcalc.org/ for details."
> Is the step-by-step usable you think? I think not. It's a sales gimmick in
> my opinion. It's not even described anywere when S-B-S works and when it
> doesn't. I have found these commands to work with
> S-B-S on the '49 (S-B-S is on, VX=X:):
<snip>
> u=X-2" "Risch alg. of tower, {'LN(X)' 'EXP(1/X)' 'X'}" "No closed form,
> LN(X)*EXP(1/X)" 'INT(LN(Xt)*EXP(1/Xt):Xt:X-2)'
>
> (At least the substitution is shown, but you keep using 'X' instead of 'U'
> or 'u'. Is the student familiar with the Risch algorithm? What is alg. short
> for? What is a tower, and has it got anything
> to do with that odd list shown?)
Hey, I'm not familiar with the Risch algorithm, and I expect that if I
want to be, I'll look it up for myself. Again, this isn't a calculator
for kids. I don't e.g... walk into a library, get out an advanced math
text on analysis, and complain when it assumes I already know the
epsilon-delta definition of continuity.
I'd love to see the day when TI give exposure of their algorithms, for
scrutiny or interest (as I believe will happen to the HP ROM within
the year)...
>
> - 'SQRT(X^2-1)' INTVX -> "SQRT(X^2-1), Square root, SQRT(X^2-1)" "Rational
> fraction, 0" "Rational fraction, (-1/2)*(2/-(2*X))"
> 'X/2*SQRT(X^2-1)+1/2*LN(SQRT(X^2-1)-X)' (Square root was surely
>
> pointed out. Rational fraction, 0? What does that mean? And where did
I agree that SBS is unclear about what algorithm it's using sometimes
(I've hardly used it, a calculator is no substitute for thinking, just
calculating)...
> error? Is the student familiar with the term 'unary operator'?
... but it's not the job of the calculator or its manual to explain
the meaning of generic terms. That's like the manual explaining,
"multiplication of a number x by a natural number y is equivalent to
adding x to itself y times". We're not trying to prove Turing
equivalence. :-)
> - [[-5 -2][2 0]] INV -> "L2=5*L2--2*L1, [[-5 -2 1 0][2 0 0 1]]" "L1=2*L1-L2,
> [[-5 -2 1 0][0 -4 2 5]]" "Reduction result, [[-10 0 0 -5][0 -4 2 5]]" [[0
> '1/2']['-1/2' '-5/4']] (Is the student
>
> familiar with the 2x4 matrices suddenly used in the calculations? What is
> the Ln representation?)
I'd hope so, yes. These look like elementary row operations, original
matrix on the left, identity on the right. I'm completely guessing
that "L" is "ligne" is French for "row", but "Line" would do.
> [[4 -7][0 35]]" "Reduction result, [[20 0][0 35]]" {[1 0] [0 1]} (I'm
> starting to see the picture. Is this much experimenting
> necessary? Is it expected by an average student?)
Well, I hope so. A calculator manual isn't designed to teach you
maths.
> Integer operations (IABCUV, ICHINREM, IDIV2, IEGCD, IQUOT & IREMAINDER):
I'm not quite sure what you're expecting. I agree that the "_________"
output is weird, but are you expecting a detailed explanation of every
message and every algorithm for which SBS is available? Are you
looking at a number theory text in trying to understand these
messages, or expecting to understand the algorithm from the calculator
output alone?
> In my opinion, PARTFRAC is useless with s-b-s mode enabled. I have written
> this to you since the first ROMs, and I know you disagree. I'd be glad to
> know what everybody else thinks?
Surprised me to see that, yes.
> The performance and novelty of the EQW is superb. That's also pretty much
> the only feature I can find, that is special and working flawlessly. That,
> and maybe MASD (almost flawless). The GUI on the '49 is also pretty fast.
Do you mean special when compared to other HP calculators? If so, the
large flash memory is nice, I'm told it's faster, and the software has
more features. The keys suck, but then I have an IBM PC XT keyboard
with nicer keys than most current keyboards -- it doesn't mean I'm
using the XT. If only I could transplant. Maybe the 49 not worth an
upgrade from a 48, but I never owned one so I didn't have to make that
painful decision.
> > And the fact that people continue to ask whether they should buy
> > a TI or an HP is a sign that it is in the competition.
>
> Yes, because the people asking trust the papers. When they try the real
> thing, they are rarely satisfied. Haven't you noticed that?
I agree. In my case, I was not satisfied by TI, so went for HP. The TI
software felt built for high school. Which is great if that's where
you are. TI was going to make it hard as hell for me to develop for or
tweak its machine, its community seemed to be 5% bright people and 95%
WHERE CAN I GET GAMEZ FOR MY TI!?!, it informed my mind how I like CAS
results presented, and [ insert usual religious pro-RPN rant here ].
And I don't think people are swayed easily into buying HP. TI are more
widely available, more widely advertised, and "your friend" is more
likely to have one. It a lot more (worthwhile) effort to choose HP
than TI.
> I still need to see the advantage of current Linux implementations. As I see
> it, Linux has its force in simplicity. Many implementations today are still
> bloated. Where I work, we use Linux for many
> purposes, mainly because it's free. Where we have mission critical systems,
> we do not use Linux (or Windows for that matter).
You wouldn't put Red Hat 7.2 Workstation Install on a PDA, nor on a
"mission critical server" (arguably, you wouldn't use Red Hat at all
on a mission critical server, but a more conservative, stable
distribution). The fact that Linux *can* be bloated out of simplicity
doesn't make it bad.
> For CPUs that are
> interesting in current handheld devices, there still
> exist many specialized OS'es. They are expensive, but very stable kernels.
Usually to make up for an extremely low horsepower CPU, as with
PalmOS, or HP49's OS (or lack of). I cannot think of a reason *not* to
use a Linux or BSD derivative (choose your licensing poison) once you
move into, say, ARM territory -- the advantages of a mature,
widely-deployed OS far outweigh any magic you'll get from a
little-used custom system.
> C/C++/EC++ compilers also exist for almost every CPU nowadays, so there is
> not much need for Java.
Agreed completely. Except to be fad-compliant there would have to be a
dotNET calculator :-). *shudder*
-- Tom
Jean-Yves Avenard
"Timité Hassan" <timi...@yahoo.fr> wrote in message
news:80e7cd6f.02031...@posting.google.com...
> What the hell does that mean ?
> You say that the HP49 has been correctly funded ?
> Then why the Hell have you needed to design and to sell the HP48G+ ?
> Why have you wasted some time and some money on this calculator when
> you could have directly worked on the HP49 ?
Do you realize the work involved to create the HP48G+ ? Almost zero. Just
change the internal memory chip and redo the faceplate.
Even today you still have people buying the HP48G+, it's cheap and has a lot
of memory. It's a different product and that way HP increased its product
range at a minimal cost and effort. So why not do it ?
There was no waste here, just pure profit. The HP48G+ has been way more
succesful in term of sales than the HP48GX in its entire sales history.
Jean-Yves
Steen Schmidt
of
> > now, the HP49G does not have a stable release ROM.
> How can you argue against the quality of a product because of its
> name? ("beta", whatever) That's classic judgement of a book by its
> cover.
Most act that way. You'd need very good inside knowledge to know if a given
beta was stable or not. There is a reason it haven't got released, and
that's because it haven't passed proper testing yet. That measn that it's
entirely guesswork if you mean that v1.1.9.6 is stable.
> To decide which software to install, on all my systems from servers to
> calculators, I observe the actions of expert users and do my own
> evaluation.
Does everybody do that? I do not hope that you are responsible for any
mission critical equipment, gievn your lack of concern about using beta
software.
> I don't blindly trust a company. Two days after buying my
> HP49, I installed 1.19.6. It was completely painless.
Were you able to determine in two days, that ROM v1.1.9.6 was stable? You're
one in a million, man.
> Do you install Microsoft's latest OS the day of its gold release,
> because Microsoft says it's release quality? Do you apply exactly the
> patches it tells you, when it tells you? And if you do, is it because
> they will provide support and pay damages if it goes wrong? (Ha!)
I can see you're a Linux guy, and colored at that too.
> While corporate incompetence may have affected technical quality, I'm
> not going to pre-judge the technical quality *because of that
> incompetence*. As far as I'm concerned, the ideas and overall design
> of the HP calculator software are excellent
There is simple not one single bug in the '49 software, when it comes to
v1.1.9.6? I can see that you have looked thoroughly indeed. What about the
flaw in list processing? Try doing { 15 90 50 } 5 /. That yields { 3. 18.
10. } in exact mode - not really that good, huh? And that one cannot be
fixed, since it's embedded beyond repair into the system.
> I switched from TI-89 to HP49, and found the former marginally faster
> for some operations. Certainly not enough of an improvement to
> influence which one I use.
You're not using many common math function then, I pressume?
> I agree that *no* calculator has a state of
> the art processor. For those moments when your calculator is a
> replacement for Mathematica/Mathcad/whatever, that's relevant. For
> now, when I need that power, on goes the laptop.
The '49 is unnecessarily ancient. That's really ok, by itself, but when
someone suddenly clame that the '49 is the best thing since sliced bread,
there's a HP zealot for ya!
> An ARM-based calculator would be nice instead for on-board power, but
> not if I have to recharge at the day's end.
How about every other month? You're very skeptical about the ARM cores, if
you think such a device necessarily needs to be recharged every day.
> I know ARM assembler, but
> I'm not sure if I want to learn another :-).
Another what? Any device based on any current architecture, one should be
able to program in a higher level language such as C or C++.
> 2. I downloaded the AUG and read most of it cover to cover, excluding
> the command reference, which I use as a reference :-).
So, can you tell me the how SOLVE works, and what its limitations are?
What's the maximum order you can develop a series to? Why? Why doesn't
PSI(1,n), where n is equal, get evaluated in exact mode? How come GAMMA(1/2)
gets evaluated in exact mode, but GAMMA(1/4) doesn't? Why does 'LN(X)' 'X=2'
SUBST demand RAD mode?
> 3. I downloaded various other docs from hpcalc.org, e.g. a UserRPL
> tutorial, which I worked through during spare moments. I grabbed some
> other docs that I will/have read when I'm ready, such as a SysRPL
> manual.
And all that in the couple of days it took you to decide to use ROM v1.19.6?
Bravo!
> 4. I read the newsgroup. From it and other documents, I noted some
> sexy apps to install, such as Emacs, which made my life easier.
Because you already code in SysRPL, or...?
> In fact, it's not dissimilar to the way I learn how to use FreeBSD or
> Linux. And that's exactly what I wanted.. a few helpful and
> enthusiastic people in a community are infinitely better than one huge
> company which gives you reams of faceless documentation then leaves
> you out in the cold (Microsoft and TI are prime examples of this).
LOL - when have you last seen JYA or BP give insights to the '49? HP is just
such a company.
> I find it hard to grasp that anyone who needs the power of an HP49
> can't find adequate information, assuming Internet access. The HP49
> isn't a lazy person's calculator, nor a high school calculator (no
> matter what HP tried to make it).
But nonetheless, that was what it was sold as. So, students should have
realized that the '49 wasn't for them, although HP told them that it was?
> How about this on the first page of the manual: "Most documentation is
> provided on line. See http://www.hpcalc.org/ for details."
How about HP provided the information, instead of leaching of off a private
website. With this community, I doubt HP would have sold even a small
fraction of the '49s they did.
> Hey, I'm not familiar with the Risch algorithm, and I expect that if I
> want to be, I'll look it up for myself.
Because it's that simple to understand, right? Get real. Does the box say;
"The HP49G can show you the steps of all calculations made, but it requires
a broad knowledge of calculus and French ways of teaching. You need to know
for example the Risch algorithm, any applicable heuristics and all other
algorithms normally used".
Who needs step-by-step capabilities, if they already know how to do it?
> Again, this isn't a calculator
> for kids.
But it is sold as one.
> I don't e.g... walk into a library, get out an advanced math
> text on analysis, and complain when it assumes I already know the
> epsilon-delta definition of continuity.
But the Librarians wouldn't guide you to that book, given your current
knowledge then. HP would.
> I'd love to see the day when TI give exposure of their algorithms, for
> scrutiny or interest (as I believe will happen to the HP ROM within
> the year)...
TI has disclosed many if not all of their algorithms used in the TI89/92
series.
> I agree that SBS is unclear about what algorithm it's using sometimes
> (I've hardly used it, a calculator is no substitute for thinking, just
> calculating)...
SBS is unclear.......well, the sole use of SBS is to shed some light on
something that might already be unclear. SBS on the '49 does not do that.
> > error? Is the student familiar with the term 'unary operator'?
>
> ... but it's not the job of the calculator or its manual to explain
> the meaning of generic terms.
What's a generic term then? Why explain differentiation? Why explain
differential equations? Why explain partial fraction decomposition? Why
explain basis of a vector space? Why explain image of a linear application
of a matrix? Why have HELP at all, and why even have a manual? With a decent
naming convention of the commands, all should be clear?
> That's like the manual explaining,
> "multiplication of a number x by a natural number y is equivalent to
> adding x to itself y times". We're not trying to prove Turing
> equivalence.
Are as many familiar with the Risch algorithm, as are familiar with
multiplication? That's not a fair comparison.
> > familiar with the 2x4 matrices suddenly used in the calculations? What
is
> > the Ln representation?)
> I'd hope so, yes. These look like elementary row operations, original
> matrix on the left, identity on the right.
And everyone buying a HP49G should be aware of this. If not, they sure can't
look for help in the manual, hoping to learn it. They need a teacher. You
have not seen older HP manuals I gather - they must have been very bad ones,
as they try to explain entirely too much.
> I'm completely guessing
> that "L" is "ligne" is French for "row", but "Line" would do.
Why not R for row?
> > [[4 -7][0 35]]" "Reduction result, [[20 0][0 35]]" {[1 0] [0 1]} (I'm
> > starting to see the picture. Is this much experimenting
> > necessary? Is it expected by an average student?)
>
> Well, I hope so. A calculator manual isn't designed to teach you
> maths.
It once was, with HP. Is a calculator an aid to teach you math? If not, then
you wouldn't really need one, as you're required to show your work step by
step when delivering answers at an exam.
> I'm not quite sure what you're expecting. I agree that the "_________"
> output is weird, but are you expecting a detailed explanation of every
> message and every algorithm for which SBS is available?
Yes, or else SBS isn't worth much.
> Are you
> looking at a number theory text in trying to understand these
> messages, or expecting to understand the algorithm from the calculator
> output alone?
Both, if possible.
What do you expect SBS to show?
> Do you mean special when compared to other HP calculators?
Special in context with calculators (or small handheld devices really) in
general.
> If so, the
> large flash memory is nice,
TI has had Flash memory for years. It's not new in a handheld device. Think
about what had happened if the '49 didn't have Flash?
> I'm told it's faster,
Faster than what? A HP48G? The '49 is slower in 99% of all cases, really.
> and the software has
> more features.
Had Erable, MetaKernel, Alg48 and MASD been subject to the same efforts,
there wouldn't really be a difference. I do not believe the features of the
'49 are finished - there are still bugs and inconsistencies. They will newer
get fixed. Have a look in the ROM, and you'll know why.
> The keys suck, but then I have an IBM PC XT keyboard
> with nicer keys than most current keyboards -- it doesn't mean I'm
> using the XT.
Why not?
> Maybe the 49 not worth an
> upgrade from a 48, but I never owned one so I didn't have to make that
> painful decision.
Or comparison! If you're used to TI, fine, but you're missing big points.
> And I don't think people are swayed easily into buying HP. TI are more
> widely available, more widely advertised, and "your friend" is more
> likely to have one. It a lot more (worthwhile) effort to choose HP
> than TI.
Because you'll have more problems using it? That doesn't seem sane.
> You wouldn't put Red Hat 7.2 Workstation Install on a PDA, nor on a
> "mission critical server" (arguably, you wouldn't use Red Hat at all
> on a mission critical server, but a more conservative, stable
> distribution).
No, I wouldn't. Would you be happy flying an F16 with Linux running your
Missile Warner, Jammer and Chaff/Flare dispensers? I wouldn't, hence we
don't use it for mission critical projects.
> The fact that Linux *can* be bloated out of simplicity
> doesn't make it bad.
It does in all the boated cases. Everyone hears about Linux, and hears that
it's good. 99% of all project managers and system administrators doesn't
have a clue about Liniux, and hence they should have kept Windows. Linux is
not easy to set up, but of course it does have its advantages. Not many
*can* take advantage of Linux.
> > For CPUs that are
> > interesting in current handheld devices, there still
> > exist many specialized OS'es. They are expensive, but very stable
kernels.
> Usually to make up for an extremely low horsepower CPU, as with
> PalmOS, or HP49's OS (or lack of).
I'm not thinking about those toys, but what about current ARM and Hitachi
implementatiotns?
> I cannot think of a reason *not* to
> use a Linux or BSD derivative (choose your licensing poison) once you
> move into, say, ARM territory -- the advantages of a mature,
> widely-deployed OS far outweigh any magic you'll get from a
> little-used custom system.
Little used is a stretch, but Linux may be wider accepted. Why would I use
50-100 times the horsepower to run a Linux kernel, when all I need is a
ARM-asm/Thumb kernel of a kilobyte or two? Decent Linux OS'es for ARM cores
aren't actually free either, so the major advantage here disappears. Most
other OS'es for ARM like cores you buy without royalties, but not many Linux
OS distributors allow that.
I'm not saying that Linux isn't a possibility in the handheld world, I think
it is, and a good one many times. That's not the same as I can make better
use of it than for example a specialized OS. That question is moot though,
as we're coding our own OS.
Regards
Steen
parisse
> beta was stable or not. There is a reason it haven't got released, and
> that's because it haven't passed proper testing yet. That measn that it's
> entirely guesswork if you mean that v1.1.9.6 is stable.
>
I don't think anyone reading comp.sys.hp48 would doubt that 1.19.6
is more stable than 1.18. You don't need to be an expert, it's just
whether you have more confidence in all the users in this forum
(including the developpers) or in the people who decide to label
a RO
M.
> Does everybody do that? I do not hope that you are responsible for any
> mission critical equipment, gievn your lack of concern about using beta
> software.
>
I generally tend to follow the developper advice when using their
software.
>
> So, can you tell me the how SOLVE works, and what its limitations are?
> What's the maximum order you can develop a series to? Why? Why doesn't
> PSI(1,n), where n is equal, get evaluated in exact mode? How come GAMMA(1/2)
> gets evaluated in exact mode, but GAMMA(1/4) doesn't? Why does 'LN(X)' 'X=2'
> SUBST demand RAD mode?
>
The kind of documentation you would require would cover 2 books
of the size of the 48 AUG. I don't think it would have been worth
the effort (and we would anyway not have time to do it). Does it
mean we should not have programmed the feature? Should I have
stopped all work on the 49 in 1999 just because HP would not release
any stable ROM release?
Usual features are documented either in HP documentation
or in Renee's documentation (which was available in French from
the beginning and was distributed by HP France in 1999)
> LOL - when have you last seen JYA or BP give insights to the '49? HP is just
> such a company.
>
As Jean-Yves said, the CAS will be released under the LGPL license.
That means that starting mid-2003, you will have full access
to the CAS source and you will be able to make your own CAS. I don't
think you will ever be allowed to read TI source code.
> Because it's that simple to understand, right? Get real. Does the box say;
> "The HP49G can show you the steps of all calculations made, but it requires
> a broad knowledge of calculus and French ways of teaching. You need to know
> for example the Risch algorithm, any applicable heuristics and all other
> algorithms normally used".
>
You don't need to know the Risch algorithm to understand SBS
integration.
You need to know usual methods. Risch is displayed to explain you
that when the usual implemented methods fail, the calc will use this
algorithm.
> TI has disclosed many if not all of their algorithms used in the TI89/92
> series.
>
I would be very interested to know where. The TI SDK does not provide
much information about CAS algorithm.
>
> Why not R for row?
>
Because L is the usual letter in French and it can be understood as
line in English.
> Decent Linux OS'es for ARM cores
> aren't actually free either, so the major advantage here disappears. Most
> other OS'es for ARM like cores you buy without royalties, but not many Linux
> OS distributors allow that.
>
Last time I heard about, the Familiar and Intimate ARM-distribution
were free (libre).
Bernard Parisse
P.S.: I hope you will change somehow your opinion when you will
have advanced in your project. Remember that we had less than 1 year
to make the 49 and we had to keep 48 compatibility. Is it better
to do nothing just because you know it will not be perfect?
parisse
> beta was stable or not. There is a reason it haven't got released, and
> that's because it haven't passed proper testing yet. That measn that it's
> entirely guesswork if you mean that v1.1.9.6 is stable.
>
I don't think anyone reading comp.sys.hp48 would doubt that 1.19.6
is more stable than 1.18. You don't need to be an expert, it's just
whether you have more confidence in all the users in this forum
(including the developpers) or in the people who decide to label
a RO
M.
> Does everybody do that? I do not hope that you are responsible for any
> mission critical equipment, gievn your lack of concern about using beta
> software.
>
I generally tend to follow the developper advice when using their
software.
>
> So, can you tell me the how SOLVE works, and what its limitations are?
> What's the maximum order you can develop a series to? Why? Why doesn't
> PSI(1,n), where n is equal, get evaluated in exact mode? How come GAMMA(1/2)
> gets evaluated in exact mode, but GAMMA(1/4) doesn't? Why does 'LN(X)' 'X=2'
> SUBST demand RAD mode?
>
The kind of documentation you would require would cover 2 books
of the size of the 48 AUG. I don't think it would have been worth
the effort (and we would anyway not have time to do it). Does it
mean we should not have programmed the feature? Should I have
stopped all work on the 49 in 1999 just because HP would not release
any stable ROM release?
Usual features are documented either in HP documentation
or in Renee's documentation (which was available in French from
the beginning and was distributed by HP France in 1999)
> LOL - when have you last seen JYA or BP give insights to the '49? HP is just
> such a company.
>
As Jean-Yves said, the CAS will be released under the LGPL license.
That means that starting mid-2003, you will have full access
to the CAS source and you will be able to make your own CAS. I don't
think you will ever be allowed to read TI source code.
> Because it's that simple to understand, right? Get real. Does the box say;
> "The HP49G can show you the steps of all calculations made, but it requires
> a broad knowledge of calculus and French ways of teaching. You need to know
> for example the Risch algorithm, any applicable heuristics and all other
> algorithms normally used".
>
You don't need to know the Risch algorithm to understand SBS
integration.
You need to know usual methods. Risch is displayed to explain you
that when the usual implemented methods fail, the calc will use this
algorithm.
> TI has disclosed many if not all of their algorithms used in the TI89/92
> series.
>
I would be very interested to know where. The TI SDK does not provide
much information about CAS algorithm.
>
> Why not R for row?
>
Because L is the usual letter in French and it can be understood as
line in English.
> Decent Linux OS'es for ARM cores
> aren't actually free either, so the major advantage here disappears. Most
> other OS'es for ARM like cores you buy without royalties, but not many Linux
> OS distributors allow that.
>
Last time I heard about, the Familiar and Intimate ARM-distribution
parisse
> beta was stable or not. There is a reason it haven't got released, and
> that's because it haven't passed proper testing yet. That measn that it's
> entirely guesswork if you mean that v1.1.9.6 is stable.
>
I don't think anyone reading comp.sys.hp48 would doubt that 1.19.6
is more stable than 1.18. You don't need to be an expert, it's just
whether you have more confidence in all the users in this forum
(including the developpers) or in the people who decide to label
a RO
M.
> Does everybody do that? I do not hope that you are responsible for any
> mission critical equipment, gievn your lack of concern about using beta
> software.
>
I generally tend to follow the developper advice when using their
software.
>
> So, can you tell me the how SOLVE works, and what its limitations are?
> What's the maximum order you can develop a series to? Why? Why doesn't
> PSI(1,n), where n is equal, get evaluated in exact mode? How come GAMMA(1/2)
> gets evaluated in exact mode, but GAMMA(1/4) doesn't? Why does 'LN(X)' 'X=2'
> SUBST demand RAD mode?
>
The kind of documentation you would require would cover 2 books
of the size of the 48 AUG. I don't think it would have been worth
the effort (and we would anyway not have time to do it). Does it
mean we should not have programmed the feature? Should I have
stopped all work on the 49 in 1999 just because HP would not release
any stable ROM release?
Usual features are documented either in HP documentation
or in Renee's documentation (which was available in French from
the beginning and was distributed by HP France in 1999)
> LOL - when have you last seen JYA or BP give insights to the '49? HP is just
> such a company.
>
As Jean-Yves said, the CAS will be released under the LGPL license.
That means that starting mid-2003, you will have full access
to the CAS source and you will be able to make your own CAS. I don't
think you will ever be allowed to read TI source code.
> Because it's that simple to understand, right? Get real. Does the box say;
> "The HP49G can show you the steps of all calculations made, but it requires
> a broad knowledge of calculus and French ways of teaching. You need to know
> for example the Risch algorithm, any applicable heuristics and all other
> algorithms normally used".
>
You don't need to know the Risch algorithm to understand SBS
integration.
You need to know usual methods. Risch is displayed to explain you
that when the usual implemented methods fail, the calc will use this
algorithm.
> TI has disclosed many if not all of their algorithms used in the TI89/92
> series.
>
I would be very interested to know where. The TI SDK does not provide
much information about CAS algorithm.
>
> Why not R for row?
>
Because L is the usual letter in French and it can be understood as
line in English.
> Decent Linux OS'es for ARM cores
> aren't actually free either, so the major advantage here disappears. Most
> other OS'es for ARM like cores you buy without royalties, but not many Linux
> OS distributors allow that.
>
Last time I heard about, the Familiar and Intimate ARM-distribution
parisse
> beta was stable or not. There is a reason it haven't got released, and
> that's because it haven't passed proper testing yet. That measn that it's
> entirely guesswork if you mean that v1.1.9.6 is stable.
>
I don't think anyone reading comp.sys.hp48 would doubt that 1.19.6
is more stable than 1.18. You don't need to be an expert, it's just
whether you have more confidence in all the users in this forum
(including the developpers) or in the people who decide to label
a RO
M.
> Does everybody do that? I do not hope that you are responsible for any
> mission critical equipment, gievn your lack of concern about using beta
> software.
>
I generally tend to follow the developper advice when using their
software.
>
> So, can you tell me the how SOLVE works, and what its limitations are?
> What's the maximum order you can develop a series to? Why? Why doesn't
> PSI(1,n), where n is equal, get evaluated in exact mode? How come GAMMA(1/2)
> gets evaluated in exact mode, but GAMMA(1/4) doesn't? Why does 'LN(X)' 'X=2'
> SUBST demand RAD mode?
>
The kind of documentation you would require would cover 2 books
of the size of the 48 AUG. I don't think it would have been worth
the effort (and we would anyway not have time to do it). Does it
mean we should not have programmed the feature? Should I have
stopped all work on the 49 in 1999 just because HP would not release
any stable ROM release?
Usual features are documented either in HP documentation
or in Renee's documentation (which was available in French from
the beginning and was distributed by HP France in 1999)
> LOL - when have you last seen JYA or BP give insights to the '49? HP is just
> such a company.
>
As Jean-Yves said, the CAS will be released under the LGPL license.
That means that starting mid-2003, you will have full access
to the CAS source and you will be able to make your own CAS. I don't
think you will ever be allowed to read TI source code.
> Because it's that simple to understand, right? Get real. Does the box say;
> "The HP49G can show you the steps of all calculations made, but it requires
> a broad knowledge of calculus and French ways of teaching. You need to know
> for example the Risch algorithm, any applicable heuristics and all other
> algorithms normally used".
>
You don't need to know the Risch algorithm to understand SBS
integration.
You need to know usual methods. Risch is displayed to explain you
that when the usual implemented methods fail, the calc will use this
algorithm.
> TI has disclosed many if not all of their algorithms used in the TI89/92
> series.
>
I would be very interested to know where. The TI SDK does not provide
much information about CAS algorithm.
>
> Why not R for row?
>
Because L is the usual letter in French and it can be understood as
line in English.
> Decent Linux OS'es for ARM cores
> aren't actually free either, so the major advantage here disappears. Most
> other OS'es for ARM like cores you buy without royalties, but not many Linux
> OS distributors allow that.
>
Last time I heard about, the Familiar and Intimate ARM-distribution
parisse
> beta was stable or not. There is a reason it haven't got released, and
> that's because it haven't passed proper testing yet. That measn that it's
> entirely guesswork if you mean that v1.1.9.6 is stable.
>
I don't think anyone reading comp.sys.hp48 would doubt that 1.19.6
is more stable than 1.18. You don't need to be an expert, it's just
whether you have more confidence in all the users in this forum
(including the developpers) or in the people who decide to label
a RO
M.
> Does everybody do that? I do not hope that you are responsible for any
> mission critical equipment, gievn your lack of concern about using beta
> software.
>
I generally tend to follow the developper advice when using their
software.
>
> So, can you tell me the how SOLVE works, and what its limitations are?
> What's the maximum order you can develop a series to? Why? Why doesn't
> PSI(1,n), where n is equal, get evaluated in exact mode? How come GAMMA(1/2)
> gets evaluated in exact mode, but GAMMA(1/4) doesn't? Why does 'LN(X)' 'X=2'
> SUBST demand RAD mode?
>
The kind of documentation you would require would cover 2 books
of the size of the 48 AUG. I don't think it would have been worth
the effort (and we would anyway not have time to do it). Does it
mean we should not have programmed the feature? Should I have
stopped all work on the 49 in 1999 just because HP would not release
any stable ROM release?
Usual features are documented either in HP documentation
or in Renee's documentation (which was available in French from
the beginning and was distributed by HP France in 1999)
> LOL - when have you last seen JYA or BP give insights to the '49? HP is just
> such a company.
>
As Jean-Yves said, the CAS will be released under the LGPL license.
That means that starting mid-2003, you will have full access
to the CAS source and you will be able to make your own CAS. I don't
think you will ever be allowed to read TI source code.
> Because it's that simple to understand, right? Get real. Does the box say;
> "The HP49G can show you the steps of all calculations made, but it requires
> a broad knowledge of calculus and French ways of teaching. You need to know
> for example the Risch algorithm, any applicable heuristics and all other
> algorithms normally used".
>
You don't need to know the Risch algorithm to understand SBS
integration.
You need to know usual methods. Risch is displayed to explain you
that when the usual implemented methods fail, the calc will use this
algorithm.
> TI has disclosed many if not all of their algorithms used in the TI89/92
> series.
>
I would be very interested to know where. The TI SDK does not provide
much information about CAS algorithm.
>
> Why not R for row?
>
Because L is the usual letter in French and it can be understood as
line in English.
> Decent Linux OS'es for ARM cores
> aren't actually free either, so the major advantage here disappears. Most
> other OS'es for ARM like cores you buy without royalties, but not many Linux
> OS distributors allow that.
>
Last time I heard about, the Familiar and Intimate ARM-distribution
Steen Schmidt
> is more stable than 1.18.
No, but most HP49G users aren't reading this group. Maybe 1% does, so it
doesn't matter much what's said here. It matters what HP writes on their
homepage.
> You don't need to be an expert, it's just
> whether you have more confidence in all the users in this forum
> (including the developpers) or in the people who decide to label
> a ROM.
A typical user trust HP, and HP does not want users to use beta ROMs.
> I generally tend to follow the developper advice when using their
> software.
You mean the company, right? HP does not support beta ROMs. If HP supported
v1.19.6, it would be fine, but HP does not.
> The kind of documentation you would require would cover 2 books
> of the size of the 48 AUG. I don't think it would have been worth
> the effort (and we would anyway not have time to do it).
Why wouldn't it be worth the effort? Aren't you interested in as many people
liking your product? Do you remember how many of the complaints about the
CAS stemmed from the fact that there was no documentation? Almost all of
them I think.
> Does it mean we should not have programmed the feature?
The importance of the manuals shouldn't be neglected. Whos idea was it to
label the '49 as a student calc by the way?
> Should I have
> stopped all work on the 49 in 1999 just because HP would not release
> any stable ROM release?
If you knew, I would have in your place.
> Usual features are documented either in HP documentation
> or in Renee's documentation (which was available in French from
> the beginning and was distributed by HP France in 1999)
French documentation is worth exactly nothing! How many of the HP49G users
know French? I would never be able to use such documentation, and I doubt
many will. Why tone the '49 so much toward the French way? I have never
understood that. The CAS, the docs - everything. Why not English, as
everyone expects?
> As Jean-Yves said, the CAS will be released under the LGPL license.
> That means that starting mid-2003, you will have full access
> to the CAS source and you will be able to make your own CAS.
I've known that for a very long time, but that doesn't matter. Are you going
to finish the '49? When? By year 2005? Can't you see; you used 1 year to
finish the hardware (with a couple of bugs), and it'll be at least 4 years
before the ROM is decent. You might as well have used 3 years to make and
good product - the saved time is your major excuse to have started the '49
project. You didn't save any time - surprised?
> I don't
> think you will ever be allowed to read TI source code.
It's not iportant. Why are you suddenly so focused on comparison with TI?
You never have before, but are we down to that level now?
> You don't need to know the Risch algorithm to understand SBS
> integration.
Oh, why not simply write "Risch algorithm used." instead of "Risch alg. of
tower, {'LN(X)' 'EXP(1/X)' 'X'}"? I do not understand the latter - do you
know anyone who does, besides you?
> You need to know usual methods.
Usual methods? I don't think many recognises the SBS messages from school. I
surely don't, anyway.
> Risch is displayed to explain you
> that when the usual implemented methods fail, the calc will use this
> algorithm.
Why try to explain the arguments you're using in that algorithm? That's no
help at all. The essence of what I'm saying about SBS, is that it currently
does only add to the confusion, it does not help the student understand how
the answer can be derived by hand.
> I would be very interested to know where. The TI SDK does not provide
> much information about CAS algorithm.
Read the numerous FAQs - many algorithms are stated and explained there. One
example could be
http://education.ti.com/product/tech/92/faqs/faq20504.html
> > Why not R for row?
> Because L is the usual letter in French and it can be understood as
> line in English.
Why stick to the French? It really sticks out. I guess French is very
universal, or were you merely making a CAS *you* can use in class? Why so
much weight on the French part - can't let it go, even though the task
demands it?
I'm sorry, but it's really weird.
> Last time I heard about, the Familiar and Intimate ARM-distribution
> were free (libre).
Familiar, what can I say. It's an unprofessional group of, as they put it,
"loosely knit people". There are still problems, even with Debian ARM
programs. Why? Because they are not consistent. The Debian ARM distribution
is not free either, but is necessary if you need decent drivers.
Intimate - this is the way to go, if you want to base yourself on the
Familiar development. It's good, it has thorough Debian ARM support, but it
weighs in at 140MB for the base image. 140MB! How on earth should that fit
in any current handheld? There are still unfixed known bugs, and not all are
found. Will they ever, with Linux?
> P.S.: I hope you will change somehow your opinion when you will
> have advanced in your project. Remember that we had less than 1 year
> to make the 49 and we had to keep 48 compatibility. Is it better
> to do nothing just because you know it will not be perfect?
You have had 3 and soon 4 years in total. In that time, you could do
practically everything, especially with that huge HP financial backing you
had.....
We have all learned from this I hope, but why fight it? Do you agree with
some here that the '49 is God almighty impersonated? I think it lacks
bigtime, and it didn't matter as long as development continued - it could
get fixed over time. Now, it won't. LGPL? Hah! You know the ROM. You, Mika,
JYA, CdB, Gerald and maybe a couple of others can edit the ROM. Nobody else
will have a snowballs chance in hell.
Regards
Steen
Colin Croft
> What about the flaw in list processing?
> Try doing { 15 90 50 } 5 /. That yields { 3. 18. 10. }
> in exact mode - not really that good, huh? And that one cannot be
> fixed, since it's embedded beyond repair into the system.
I'm fascinated... Please forgive my ignorance but I'm not a 49G user:
what's wrong with dividing {15,90,50} by 5 and getting {3,18,10}? This
is exactly what you'd get on a 39G.
Steen Schmidt
> what's wrong with dividing {15,90,50} by 5 and getting {3,18,10}? This
> is exactly what you'd get on a 39G.
Check out that the numbers in the list after the division are always numeric
type - not exact integers. You can better see the difference with
multiplication:
{ 4 5 6 } 4 * -> { 16. 20. 24. } (numeric)
4 { 4 5 6 } * -> { 16 20 24 } (exact)
Both should yield a exact result in exact mode, when all inputs are exact.
Regards
Steen
Tom Garcia
"Steen Schmidt" <ssch...@nospam.dk> wrote in message news:<QUwj8.479$pQ.1...@news010.worldonline.dk>...
> > > If v1.19.6 was relabeled v1.20, and made a release ROM I'd be happy. As
> of
> > > now, the HP49G does not have a stable release ROM.
> > How can you argue against the quality of a product because of its
> > name? ("beta", whatever) That's classic judgement of a book by its
> > cover.
>
> Most act that way. You'd need very good inside knowledge to know if a given
> beta was stable or not.
And you need a good "inside" knowledge to know whether an official
release is stable. Who trusts the word of the average software
company? I certainly don't.
> There is a reason it haven't got released, and
> that's because it haven't passed proper testing yet.
I guess 1.0.5 was more stable then, because it wasn't marked "beta".
> > To decide which software to install, on all my systems from servers to
> > calculators, I observe the actions of expert users and do my own
> > evaluation.
>
> Does everybody do that?
Yes, I would hope so. Independent thought rather than relying on a
company's advertising generally results in your selecting a better
product.
> > I don't blindly trust a company. Two days after buying my
> > HP49, I installed 1.19.6. It was completely painless.
>
> Were you able to determine in two days, that ROM v1.1.9.6 was stable? You're
> one in a million, man.
I was able to judge that, given people's comments, it was worthwhile
trying it out. I could always go back, and hey, it was a calculator,
who cares if I get a few hours of poor performance? It was a weekend
anyway. If the move had been irreversible, I'd have taken longer.
> > Do you install Microsoft's latest OS the day of its gold release,
>
> I can see you're a Linux guy, and colored at that too.
No, I'm mostly a FreeBSD guy and a Windows guy. Whatever gets the job
done.
> There is simple not one single bug in the '49 software, when it comes to
> v1.1.9.6? I can see that you have looked thoroughly indeed.
Why would I make sure software is 100% bug free before using it? That
would mean using only the most trivial apps, or just dumping computers
entirely. I agree the 49 has some nasty bugs, but none that have
stopped me using the machine.
> > I switched from TI-89 to HP49, and found the former marginally faster
> > for some operations. Certainly not enough of an improvement to
> > influence which one I use.
>
> You're not using many common math function then, I pressume?
Ok the general TIvsHP battle isn't worth it..
> The '49 is unnecessarily ancient. That's really ok, by itself, but when
> someone suddenly clame that the '49 is the best thing since sliced bread,
> there's a HP zealot for ya!
Agreed. It just does the job.
> > An ARM-based calculator would be nice instead for on-board power, but
> > not if I have to recharge at the day's end.
>
> How about every other month? You're very skeptical about the ARM cores, if
> you think such a device necessarily needs to be recharged every day.
No, I'm not skeptical, I'm very impressed.. but it's possible to
create a high power device, like a Compaq iPaq, which needs recharging
every day, with an ARM core; or an ARM7100 core'd Psion Series 5 which
swims on 30 continuous hours between recharges. All I'm saying is.. if
a new device is released, I would like it to either balance nicely
between consumption and horsepower, OR better to have a variable clock
with a *significant* range.
> > I know ARM assembler, but
> > I'm not sure if I want to learn another :-).
>
> Another what? Any device based on any current architecture, one should be
> able to program in a higher level language such as C or C++.
I wasn't being entirely serious, hence the :-). I like ARM assembler,
that's all I was saying. Its barrel shifter and use of condition codes
are things of beauty.
> > 2. I downloaded the AUG and read most of it cover to cover, excluding
> > the command reference, which I use as a reference :-).
>
> So, can you tell me the how SOLVE works, and what its limitations are?
> What's the maximum order you can develop a series to? Why? Why doesn't
<snip>
Why are you throwing bug reports at me? If your point is that
occasional bugs make it hard for the new user to learn, then yes, this
is the case with all software. Fingers crossed, the source will at
some point be out for community fixing.
> > 3. I downloaded various other docs from hpcalc.org, e.g. a UserRPL
> > tutorial, which I worked through during spare moments. I grabbed some
> > other docs that I will/have read when I'm ready, such as a SysRPL
> > manual.
>
> And all that in the couple of days it took you to decide to use ROM v1.19.6?
> Bravo!
So I spent a weekend plus an evening learning how to use a calculator,
and getting more docs to learn more. I guess I must be sad. Oh well.
> > 4. I read the newsgroup. From it and other documents, I noted some
> > sexy apps to install, such as Emacs, which made my life easier.
>
> Because you already code in SysRPL, or...?
Because it had command completion, was a reason alone to get it.
> > Linux. And that's exactly what I wanted.. a few helpful and
> > enthusiastic people in a community are infinitely better than one huge
>
> LOL - when have you last seen JYA or BP give insights to the '49? HP is just
> such a company.
I'm talking about *everyone*, not just the official developers (and
JYA seemed to give a lot more insight while working than the TI
workers did). Random example... ticalc.org "TIcalc has an IRC help
channel on EFnet, #ticalc" (or whatever it's called, can't remember
now). It doesn't. You go in there, talk about TI, and are told YOU ARE
NOT ELITE and kicked. Grow up people!
> > isn't a lazy person's calculator, nor a high school calculator (no
> > matter what HP tried to make it).
>
> But nonetheless, that was what it was sold as. So, students should have
> realized that the '49 wasn't for them, although HP told them that it was?
No, students who weren't willing to put in extra effort should have
realised that they shouldn't be buying a complex calculator *at all*.
Or if they do, then they should ensure they buy from a company with a
good returns policy.
> > How about this on the first page of the manual: "Most documentation is
> > provided on line. See http://www.hpcalc.org/ for details."
>
> How about HP provided the information, instead of leaching of off a private
> website. With this community, I doubt HP would have sold even a small
> fraction of the '49s they did.
Agreed, HP should have supported the web site, through hosting at
least, payment / special contacts / etc preferably. I'm not sure if
they should have *taken over* this support role completely. Firstly,
it's never good for a community to be totally at the mercy of a
company. Secondly, lots of enthusiasts give better support (imho) than
a few workers.
> Because it's that simple to understand, right? Get real. Does the box say;
> "The HP49G can show you the steps of all calculations made, but it requires
> a broad knowledge of calculus and French ways of teaching. You need to know
> for example the Risch algorithm, any applicable heuristics and all other
> algorithms normally used".
On the "broad knowledge of calculus", are you suggesting that someone
with no knowledge of calculus whatsoever should understand a step by
step explanation of a differentiation procedure? I agree, as in my
previous mail, that an indication of general algorithms used would be
helpful.
> Who needs step-by-step capabilities, if they already know how to do it?
I thought SBS was to indicate how to apply mostly-known methods to a
particular question, if you're having trouble with choosing a
particular step, or simply want to check your working.
> > Again, this isn't a calculator
> > for kids.
>
> But it is sold as one.
Yes. A stupid marketing mistake. But not a technical one.
> > I don't e.g... walk into a library, get out an advanced math
> > text on analysis, and complain when it assumes I already know the
> > epsilon-delta definition of continuity.
>
> But the Librarians wouldn't guide you to that book, given your current
> knowledge then. HP would.
I would say any bright high school kid or (university) college student
in any of the hard sciences/engrg, maths or computing should be able
to grasp an HP. That's the level I was thinking of, compared to other
makes which anyone from high school onwards should be able to grasp
quite easily.
I agree that pitching the HP on the same level as TI was silly. Part
of the problem is a general desire of people to do as little
thinking/understanding as possible. The HP marketers should have
recognised this consumer trend, and that you can't sell anything
unless it's spoon-fed.
> SBS is unclear.......well, the sole use of SBS is to shed some light on
> something that might already be unclear. SBS on the '49 does not do that.
Sometimes it does, sometimes it doesn't. Again, I see the SBS as a way
of checking my method, or hinting me which step to use next on a
method I already know.
> What's a generic term then?
One that has exactly the same meaning in a wide community outside of
the specific product. "unary operator" is certainly such a term.
> Why explain differentiation? Why explain <etc>
In a calculator manual, it's certainly unnecessary.
> of a matrix? Why have HELP at all, and why even have a manual? With a decent
> naming convention of the commands, all should be clear?
The manual needs, at most, to bridge the gap between your existing
knowledge of maths/science/eng and accessing the functionality as
implemented on the equipment. If more information will be provided, I
won't complain, but it's not necessary. I cannot emphasise it enough
-- I do not think a calculator is a teaching tool.
> Are as many familiar with the Risch algorithm, as are familiar with
> multiplication? That's not a fair comparison.
My point was.. "look it up". Surely a student's job is to research and
learn?
> > > familiar with the 2x4 matrices suddenly used in the calculations? What
> is
> > > the Ln representation?)
> > I'd hope so, yes. These look like elementary row operations, original
> > matrix on the left, identity on the right.
>
> And everyone buying a HP49G should be aware of this. If not, they sure can't
> look for help in the manual, hoping to learn it. They need a teacher. You
> have not seen older HP manuals I gather - they must have been very bad ones,
> as they try to explain entirely too much.
I'm not campaigning against a huge, lucid manual. I love huge, lucid
manuals. I'm just saying that your argument is pitched at a level that
removes completely from a student the requirement to think. Maybe I'm
just silly, and like investigating.
> > I'm completely guessing
> > that "L" is "ligne" is French for "row", but "Line" would do.
>
> Why not R for row?
"R" would be better. I was telling you my thought processes in working
out what it meant.
> It once was, with HP. Is a calculator an aid to teach you math? If not, then
> you wouldn't really need one, as you're required to show your work step by
> step when delivering answers at an exam.
Ah, OK, I see we have different premises here.. I consider a
calculator to be a grinding, hinting and a checking tool. You also
consider it (with its manual) to be a teaching tool. In which case,
many of your points are fair...
Using a calculator in an exam is, imho, firstly about manipulating
numbers, matrices or lists of data, doing boring arithmetic. It's
could also be about checking that you've worked out your hard integral
correctly, and even (with SBS) hinting the direction you should go to
get it, assuming you already understand the general method.
(of course, for my analysis/graph theory/number theory/logic etc.
exams this year, I'm not even allowed a calculator; only the numerical
methods exam allows one)
> > Are you
> > looking at a number theory text in trying to understand these
> > messages, or expecting to understand the algorithm from the calculator
> > output alone?
>
> Both, if possible.
I just can't begin to comprehend learning a method completely from the
output of a calculator. I assume when you learn most algorithms you
read and understand a proof of how they work. Do you expect the
calculator to provide that too?
> > If so, the
> > large flash memory is nice,
>
> TI has had Flash memory for years. It's not new in a handheld device. Think
> about what had happened if the '49 didn't have Flash?
But TI places such silly restrictions on your use of flash. For
example -- I can't write and install flash software without *paying*
TI for a key. And the more software I write, the more I have to pay.
The last time I asked TI when they were going to release a freeware
key: "Real soon now!" Has it happened yet?
> there wouldn't really be a difference. I do not believe the features of the
> '49 are finished - there are still bugs and inconsistencies. They will newer
> get fixed. Have a look in the ROM, and you'll know why.
If it goes GPL, it'll be up to everyone, not just the company, to fix
as much as possible. Or to re-use the ideas in a new, improved system.
No complex software is ever "finished".
> > The keys suck, but then I have an IBM PC XT keyboard
> > with nicer keys than most current keyboards -- it doesn't mean I'm
> > using the XT.
>
> Why not?
I said after, "If only I could transplant". I meant that I agree that
the keyboard of a 49 is poor, but unfortunately, I cannot use the
keyboard of a previous HP. So, I'm not going to use a less powerful
machine just to get a better keyboard.
> It a lot more (worthwhile) effort to choose HP than TI.
>
> Because you'll have more problems using it? That doesn't seem sane.
You'll have more problems *learning* to use an HP, yes. In terms of
programming the thing to do what you want, I know which I prefer.
> No, I wouldn't. Would you be happy flying an F16 with Linux running your
> Missile Warner, Jammer and Chaff/Flare dispensers? I wouldn't, hence we
> don't use it for mission critical projects.
No, I certainly wouldn't use Linux for real time work. But "mission
critical" is often used to mean "business critical", and for that I'd
happily employ a BSD variant (less likely Linux).
> > The fact that Linux *can* be bloated out of simplicity
> > doesn't make it bad.
>
> It does in all the boated cases. Everyone hears about Linux, and hears that
> it's good.
If you float around the BSD community, everyone hears that it is
baaaaaad :-).
> 99% of all project managers and system administrators doesn't
> have a clue about Liniux, and hence they should have kept Windows.
As IBM used to say, "Think!" Obviously, you don't convert your office
to Linux when your only sysadmins are MSCEs.
> I'm not thinking about those toys, but what about current ARM and Hitachi
> implementatiotns?
ARM are fantastic. I've had an ARM box since 1989, when they were
*Acorn* RISC Machines. I don't know much about Hitachi.
> > I cannot think of a reason *not* to
> > use a Linux or BSD derivative (choose your licensing poison) once you
> > move into, say, ARM territory -- the advantages of a mature,
> > widely-deployed OS far outweigh any magic you'll get from a
> > little-used custom system.
>
> Little used is a stretch, but Linux may be wider accepted. Why would I use
> 50-100 times the horsepower to run a Linux kernel, when all I need is a
> ARM-asm/Thumb kernel of a kilobyte or two?
I didn't realise the Linux kernel slowed down the average app 50-100
times :-). Seriously, have you tried running Unix variants on ARMs? I
used NetBSD/arm32 regularly back in 1996 or so, and ARMLinux in 98,
both on ARM6 (which I'm guessing is around the kind of power you'd
throw in a long-life handheld) and I didn't notice a huge speed drop
over a non-pre-emptive system.
> Decent Linux OS'es for ARM cores
> aren't actually free either, so the major advantage here disappears. Most
> other OS'es for ARM like cores you buy without royalties, but not many Linux
> OS distributors allow that.
I think the FSF would have a lot to say if a Linux OS didn't have its
source freely available. The licensing of Linux makes it illegal to
distribute a derivative and require payment for its source. Or am I
misunderstanding?
* * *
Thanks for taking the time to post all your thoughts. If you view the
calculator as a teaching aid, then many of your points are very fair.
Enjoy,
-- Tom
Timité Hassan
Hi Cyrille,
At first i want to apologize,because i think that i went a bit too far
in my previous message.
> > Then why the Hell have you needed to design and to sell the HP48G+ ?
> As far as any HP48 expert can say, there is no design needed to create a
> HP48G+, it's exacly the same product than a HP48G except for the memory
> device which is 128K instead of 32K (well, technicaly, it's the same than a
> GX minus the port connector, and the bank switcher). As for the selling
> part, well, if it looks like you can make a buck or to at minimal cost, why
> not do it? it's called extending a product range....
So the HP48G+ has been released just to do profits ?
> > Why have you wasted some time and some money on this calculator when
> > you could have directly worked on the HP49 ?
> A couple of possible answers: Because they needed some time to think about
> what they were going to do next, because they were sourcing the budget
> (these things takes time in a corporate environment), because they did not
> have the staff to do so, because they wanted the peoples to learn first how
> to deal with calculators by introducing a simpler product like the 48G+,
> because they were looking for partners, maybe, the 49 was already in the
> pipe and in design in parallel with the HP48G+, maybe because they had other
> product to introduce first that made more marketing sence, maybe they were
> trying to scope the work for the HP49, working on the schedule
>
> > Perhaps even you would have had enough time and money to design a
> > faster C.P.U and a better controller than the Yorke.
> Well, ASIC design time is in years from starting the work to first
> prototypes. the 49 was introduced barely a year after the ACO was created,
> Even with an experienced team, it would not have been possible to redesign
> the Yorke in such a small timeframe.
I thought that the A.C.O was created in 1997 ?
Anyway if you can't redesign the yorke then why design the HP49 ?
You could have just upgraded the HP48 with the new hardware memory
management,Flash technology ,integration of HP48 best application and
software optimisation.Considering the time you had to design the
HP49,it would have been a wiser,more secure move.This calculator would
have been a HP49 in the HP48 case without gizmo such as the algebraic
mode and HP-Basic.
> > Also if you were correctly funded for the HP49G,how come this
> > calculator had such a poor documentation and marketing ?
> > Because for me it is either a lack of money or a lack of competent
> > people.
> well, it could also be lake of time, lake of people all together, corporate
> heaviness, underscoping of tasks, lake of comunication between teams, bad
> (or changing) requierements.
> Also, remember that the 49 was the first product of a young division and
> team, a learning product... and they definitely made beginner mistakes...
OK but mistakes like that have strongly hurt your credibility because
never before a HP calculator has been released with such
problems(including hardware and software problems).
Sorry to say that but the HP49 seemed like a very rushed product.
And the youth of the A.C.O doesn't excuse all.
The designers team of the TI92 calculators was also a young team but
even if the first TI92 had hardware and software problems,the
documentation and the marketing were excellent.
> Regards, Cyrille
MSCHAEF.COM
Timité Hassan <timi...@yahoo.fr> wrote:
>"Cyrille de Br?isson" <cyrille_de...@hp.com> wrote in message
>news:<a6lirn$87r$1...@web1.cup.hp.com>...
>Hi Cyrille,
>
>At first i want to apologize,because i think that i went a bit too far
>in my previous message.
>
>> > Then why the Hell have you needed to design and to sell the HP48G+ ?
>> As far as any HP48 expert can say, there is no design needed to create a
>> HP48G+, it's exacly the same product than a HP48G except for the memory
>> device which is 128K instead of 32K (well, technicaly, it's the same than a
>> GX minus the port connector, and the bank switcher). As for the selling
>> part, well, if it looks like you can make a buck or to at minimal cost, why
>> not do it? it's called extending a product range....
>
>So the HP48G+ has been released just to do profits ?
That's generally what companies do: Try to make profits. More generally,
they try to create value for the shareholders that _own_ them and to
whom they are _legally_ obligated.
-Mike
--
http://www.mschaef.com
Bhuvanesh
> Hello, some of my thoughts...
>
> "Steen Schmidt" <ssch...@nospam.dk> wrote in message news:<8ici8.1166$Z25....@news000.worldonline.dk>...
>
> > If v1.19.6 was relabeled v1.20, and made a release ROM I'd be happy. As of
> > now, the HP49G does not have a stable release ROM.
>
> How can you argue against the quality of a product because of its
> name? ("beta", whatever) That's classic judgement of a book by its
> cover.
Steen isn't talking about the regulars here, who would try out betas;
he's talking about the average user.
> In fact, it's not dissimilar to the way I learn how to use FreeBSD or
> Linux. And that's exactly what I wanted.. a few helpful and
> enthusiastic people in a community are infinitely better than one huge
> company which gives you reams of faceless documentation then leaves
> you out in the cold (Microsoft and TI are prime examples of this).
What's "faceless" documentation? How is TI "[leaving] you out in the
cold"??
> Hey, I'm not familiar with the Risch algorithm, and I expect that if I
> want to be, I'll look it up for myself. Again, this isn't a calculator
> for kids. I don't e.g... walk into a library, get out an advanced math
> text on analysis, and complain when it assumes I already know the
> epsilon-delta definition of continuity.
But the HP49G *was* targeted to students! I'm somewhat familiar with
the (original) Risch algorithm, but that's because I learnt it by
myself. I cannot expect high-school students (or even 99.9% of college
students) to be familiar with it.
> I'd love to see the day when TI give exposure of their algorithms, for
> scrutiny or interest (as I believe will happen to the HP ROM within
> the year)...
I'm not sure that would be very useful. I'd rather they document the
algorithms they use in the Command Reference section in the Appendix
(the way they are doing now) and put some more of the internal
routines (e.g. polynomial gcd) in the jump table.
> > Integer operations (IABCUV, ICHINREM, IDIV2, IEGCD, IQUOT & IREMAINDER):
>
> I'm not quite sure what you're expecting. I agree that the "_________"
> output is weird, but are you expecting a detailed explanation of every
> message and every algorithm for which SBS is available?
Yes. If SBS is available at all, it should be detailed enough to be
usable.
> I agree. In my case, I was not satisfied by TI, so went for HP. The TI
> software felt built for high school.
And yet several professionals use it for advanced math ;-)
> [ insert usual religious pro-RPN rant here ].
:-) I do have RPN too, but on the TI-89/92+.
--
Bhuvanesh
Steen Schmidt
> release is stable. Who trusts the word of the average software
> company? I certainly don't.
Well, we share many points of view, but I'm trying to look at the HP49G as
what it was sold as - it was sold to students promising a TI89/92 basher.
That's a failure. I like the '49, and I find it a very cool toy. I like
investigations too. I have used many great hours hacking in this device and
several HP calculators before that. I also own a TI89, but it's a workhorse
currently, the '49 will always be the most fun of the two.
The audience the '49 was sold to, will usually only trust the ROM the
manufacturer backs. That is v1.18, which is pretty bad really.
I use v1.19.6, and find it very good. HP dropped its promise (was it really
ever one?) to ever get this calc working like old HP calcs, hence my remark
"Carly dropped the ball".
> I guess 1.0.5 was more stable then, because it wasn't marked "beta".
Erh, yeah ;-) v1.05 suck bigtime, and was probably released in panic. You
and I know that, but many of the "real" customers don't.
> Yes, I would hope so. Independent thought rather than relying on a
> company's advertising generally results in your selecting a better
> product.
I know, and that's what I generally do too. If we keep our eyes to the main
group the '49 was sold to, they are now abandoned.
> I was able to judge that, given people's comments, it was worthwhile
> trying it out. I could always go back, and hey, it was a calculator,
> who cares if I get a few hours of poor performance? It was a weekend
> anyway. If the move had been irreversible, I'd have taken longer.
Had it been a more important application, you wouldn't have dared to use a
beta (I hope). If not for anything else, then because of the risk of loosing
the ability to return to the manufacturer with any legal claims.
> Why would I make sure software is 100% bug free before using it? That
> would mean using only the most trivial apps, or just dumping computers
> entirely. I agree the 49 has some nasty bugs, but none that have
> stopped me using the machine.
What I mean is that the '49 do have some heavy bugs that could've been
avoided. That work should've been started more than 2 years ago - now it's
practically hopeless.
> All I'm saying is.. if
> a new device is released, I would like it to either balance nicely
> between consumption and horsepower, OR better to have a variable clock
> with a *significant* range.
You'll have both.
> > So, can you tell me the how SOLVE works, and what its limitations are?
> > What's the maximum order you can develop a series to? Why? Why doesn't
> <snip>
>
> Why are you throwing bug reports at me?
These are not bug reports, but I find the points important to know. You
can't find that information anywhere in the documentation, that was my
point. If you don't miss it, great, but if you do (because you are used to
HP thoroughnes) then too bad.
> Fingers crossed, the source will at
> some point be out for community fixing.
LOL - have you seen the sources? Good luck, I'm telling you. Good luck.
> Because it had command completion, was a reason alone to get it.
I like Emacs as well - its auto completion is superb. I just didn't think
you'd need it unless you were coding SysRPL.
> No, students who weren't willing to put in extra effort should have
> realised that they shouldn't be buying a complex calculator *at all*.
Hmm, the initial target group is usually very subsceptible to commercials.
They didn't have a chance in my opinion.
> Or if they do, then they should ensure they buy from a company with a
> good returns policy.
HP used to be such a company.
> On the "broad knowledge of calculus", are you suggesting that someone
> with no knowledge of calculus whatsoever should understand a step by
> step explanation of a differentiation procedure?
No, I'm just searching for a better structure in the SBS mode. A few
additional words, and better layout of the information would help
tremendeously. The information given, when you get down to it, is adequate.
It's merely very hard to decipher.
> Yes. A stupid marketing mistake. But not a technical one.
But a part of the general failure (who's General Failure?..).
> I would say any bright high school kid or (university) college student
> in any of the hard sciences/engrg, maths or computing should be able
> to grasp an HP.
And they would, those who would have chosen a HP by themselves. Now HP
pushed '49s on everybody - also those who should have sticked with a TI89.
> > What's a generic term then?
>
> One that has exactly the same meaning in a wide community outside of
> the specific product. "unary operator" is certainly such a term.
Of course, but not for a ten year old. In my circles, AN/ALQ149 is a generic
term, but I wouldn't expect everyone to know. Unary operator is not that
bad, but not that many high school kids realize that zero can be looked at
as a rational fraction.
> In a calculator manual, it's certainly unnecessary.
I always found it nice, but I also started reading them long before I had
the topics in school.
> I'm not campaigning against a huge, lucid manual. I love huge, lucid
> manuals. I'm just saying that your argument is pitched at a level that
> removes completely from a student the requirement to think. Maybe I'm
> just silly, and like investigating.
I agree with you, but HP tried to enter that thoughtless market. They
failed, in my opinion.
> Ah, OK, I see we have different premises here.. I consider a
> calculator to be a grinding, hinting and a checking tool. You also
> consider it (with its manual) to be a teaching tool. In which case,
> many of your points are fair...
JYA & BP exclaimed that the '49 was a teaching tool. We were all skeptical
then, and more now. They were obviously wrong.
> I just can't begin to comprehend learning a method completely from the
> output of a calculator. I assume when you learn most algorithms you
> read and understand a proof of how they work. Do you expect the
> calculator to provide that too?
No, but maybe examples in the manual would help.
> But TI places such silly restrictions on your use of flash. For
> example -- I can't write and install flash software without *paying*
> TI for a key. And the more software I write, the more I have to pay.
> The last time I asked TI when they were going to release a freeware
> key: "Real soon now!" Has it happened yet?
Yes. There are many other restriction too though.
> If it goes GPL, it'll be up to everyone, not just the company, to fix
> as much as possible. Or to re-use the ideas in a new, improved system.
> No complex software is ever "finished".
You'll just have to wait and see then. I doubt the '49 OS will be the next
Linux. Only a selcet group will ever grasp to modify the ROM. Werner is one
:-)
> You'll have more problems *learning* to use an HP, yes. In terms of
> programming the thing to do what you want, I know which I prefer.
Me too, but we're not the majority.
> No, I certainly wouldn't use Linux for real time work. But "mission
> critical" is often used to mean "business critical", and for that I'd
> happily employ a BSD variant (less likely Linux).
I agree there.
> As IBM used to say, "Think!" Obviously, you don't convert your office
> to Linux when your only sysadmins are MSCEs.
But some does - did anyone say HP here?....
> I didn't realise the Linux kernel slowed down the average app 50-100
> times :-).
Ok - I meant percent. Sorry. I surely wouldn't choose Linux if the only
thing I gained was half the performance.
> Seriously, have you tried running Unix variants on ARMs?
Yes. They are not entirely bad, but in the short time we used Linux, we (a
body of mine really, who is in charge of the OS) found too many anomalies if
you will. Bugs, although they were quickly fixed, they did emerge. then new
ones, then incompatibilities, then....
We were also sacrificing much space by maintaining a kernel like that in
Flash. Now we have a 90% working kernel, an almost finished (will it ever?)
TCP/IP stack, USB/Seriel protocols and so on. Development is pretty fast,
when you have a EC++ compiler, but we of course have a long way yet to go.
> on ARM6 (which I'm guessing is around the kind of power you'd
> throw in a long-life handheld)
No, if I must say, we're using Intels PXA210/250 and successors.
> I think the FSF would have a lot to say if a Linux OS didn't have its
> source freely available. The licensing of Linux makes it illegal to
> distribute a derivative and require payment for its source. Or am I
> misunderstanding?
The sources are free, of course, but they are not that useful without
support. We can of course compile them, but what happens when we discover a
bug? We can't force anyone to fix it. That uncertainty is hard to swallow.
Some projects are stable enough to handle such risks - ours aren't
currently. We need to bet on the safe horse, and that's not Linux right now.
It may be at a later stage.
Regards
Steen
Bhuvanesh
> > I would be very interested to know where. The TI SDK does not provide
> > much information about CAS algorithm.
>
> Read the numerous FAQs - many algorithms are stated and explained there. One
> example could be
>
> http://education.ti.com/product/tech/92/faqs/faq20504.html
And TI sends more information about internal algorithms if you ask for
it. For example, I got quite a bit of extra information about factor()
and isprime(), and it will be publicly available, either from Appendix
A of the guidebook (which already gives information about algorithms
for many functions) or in Doug's tip-list.
--
Bhuvanesh
parisse
thanks for the link, I didn't know that (I must confess I do
not care much now about non-free software). But it is anyway not useful
to write C code, or are there any info on the internal C functions
of the TI-AMS?
Timité Hassan
> > on ARM6 (which I'm guessing is around the kind of power you'd
> > throw in a long-life handheld)
>
> No, if I must say, we're using Intels PXA210/250 and successors.
I thought that you would use a C.P.U which can use a F.P.U unless the
Intel Xscale will support F.P.U ?
parisse
> doesn't matter much what's said here. It matters what HP writes on their
> homepage.
>
I don't control what HP does. I could only give them advices.
> > Should I have
> > stopped all work on the 49 in 1999 just because HP would not release
> > any stable ROM release?
>
> If you knew, I would have in your place.
>
That's an interesting point of view. Following it I should never have
fixed bugs for the same reasons.
> > Usual features are documented either in HP documentation
> > or in Renee's documentation (which was available in French from
> > the beginning and was distributed by HP France in 1999)
>
> French documentation is worth exactly nothing! How many of the HP49G users
> know French? I would never be able to use such documentation, and I doubt
> many will. Why tone the '49 so much toward the French way? I have never
> understood that. The CAS, the docs - everything. Why not English, as
> everyone expects?
>
Because sorry we are not fluent in English like you, hence the CAS doc
was written by Renee in French. I'm not responsible for HP decision
not to translate it in English. Why should the root langage always
be English? Don't you think it is as easy to translate from French
than from English? Anyway, the CAS doc was translated in English by
Ivan hence it is available in English.
> > As Jean-Yves said, the CAS will be released under the LGPL license.
> > That means that starting mid-2003, you will have full access
> > to the CAS source and you will be able to make your own CAS.
>
> I've known that for a very long time, but that doesn't matter. Are you going
> to finish the '49? When? By year 2005? Can't you see; you used 1 year to
> finish the hardware (with a couple of bugs), and it'll be at least 4 years
> before the ROM is decent. You might as well have used 3 years to make and
> good product - the saved time is your major excuse to have started the '49
> project. You didn't save any time - surprised?
>
I don't work for HP. The work I made after the release of the 49 was
on my free time. I have fixed all CAS-related bugs I could. I don't
have time to enhance the 49 CAS, but everyone interested will have
this opportunity in mid-2003. What else could I do?
> > I don't
> > think you will ever be allowed to read TI source code.
>
> It's not iportant. Why are you suddenly so focused on comparison with TI?
> You never have before, but are we down to that level now?
>
You raised the subject by comparing internals of both calcs.
> > You don't need to know the Risch algorithm to understand SBS
> > integration.
>
> Oh, why not simply write "Risch algorithm used." instead of "Risch alg. of
> tower, {'LN(X)' 'EXP(1/X)' 'X'}"? I do not understand the latter - do you
> know anyone who does, besides you?
>
Read for example Davenport-Siret-Tournier.
> > You need to know usual methods.
>
> Usual methods? I don't think many recognises the SBS messages from school. I
> surely don't, anyway.
>
Which shows that the way we teach in France is somewhat different
from elsewhere.
> Why stick to the French? It really sticks out. I guess French is very
> universal, or were you merely making a CAS *you* can use in class? Why so
> much weight on the French part - can't let it go, even though the task
> demands it?
>
French is not as universal as English, but it is also an
universal language as are e.g. Chinese, Spanish and many other
languages. It would be a big loss if everything would be replaced
by one language. And there is no reason that English will remain
dominant in the future.
>
> Familiar, what can I say. It's an unprofessional group of, as they put it,
> "loosely knit people". There are still problems, even with Debian ARM
> programs. Why? Because they are not consistent. The Debian ARM distribution
> is not free either, but is necessary if you need decent drivers.
>
It seems we are engaged in an endless debate because I don't believe
the world should be built following the model of the Cathedral
but instead of the Bazaar.
>
> You have had 3 and soon 4 years in total. In that time, you could do
> practically everything, especially with that huge HP financial backing you
> had.....
>
You can't speak from this subject since you certainly do not have any
information.
> We have all learned from this I hope, but why fight it? Do you agree with
> some here that the '49 is God almighty impersonated? I think it lacks
> bigtime, and it didn't matter as long as development continued - it could
> get fixed over time. Now, it won't. LGPL? Hah! You know the ROM. You, Mika,
> JYA, CdB, Gerald and maybe a couple of others can edit the ROM. Nobody else
> will have a snowballs chance in hell.
>
Why? When the CAS will be released under the LGPL, people will have
much more info and tools than we had when we started coding for the 48
and they
will be able to *modify* their OS in flash. And
there are certainly good coders in the young generation. Maybe they
will prefer coding other projects, maybe not, who knows?
Steen Schmidt
>
> I thought that you would use a C.P.U which can use a F.P.U unless the
> Intel Xscale will support F.P.U ?
As you now know, the XScale core has a DSP internally. It has advantages and
drawbacks compared to a dedicated FPU, but it'll suffice. The loss is not
that great.
Regards
Steen
Steen Schmidt
But you argued that the beta ROM issue was not a problem, and I argue it is.
You say that it is not, since everyone reading here will know to use
v1.19.6. Then I say that almost none of the users are reading c.s.h. Do you
admit that the beta ROM is indeed an issue now?
> > If you knew, I would have in your place.
> That's an interesting point of view. Following it I should never have
> fixed bugs for the same reasons.
If you *knew* HP would act like they did, would you seriously use all your
good free time on fixing bugs in the '49? I understand that you'd under all
circumstances work on Giac, but if you kept the two seperate?
Would you use your spare time, with the knowledge you have now? I wouldn't.
> Because sorry we are not fluent in English like you, hence the CAS doc
> was written by Renee in French. I'm not responsible for HP decision
> not to translate it in English.
I'm not accusing you for not writing it in English, I'm just saying that
having it in French is not the big advantage you say it is.
HPs decission not to release more and better documentation is the big bad
wolf here - I never said it was you.
> Why should the root langage always
> be English?
Because so many uses it?
> Don't you think it is as easy to translate from French
> than from English?
Not for me, and not for the biggest part of users.
> Anyway, the CAS doc was translated in English by
> Ivan hence it is available in English.
Yes, and HP should have done that a long time ago.
> I don't work for HP. The work I made after the release of the 49 was
> on my free time. I have fixed all CAS-related bugs I could. I don't
> have time to enhance the 49 CAS, but everyone interested will have
> this opportunity in mid-2003. What else could I do?
I think that you have done a brilliant job, given the circumstances - I was
merely arguing that the time it took to get this far with the whole project,
is not nay shorter than if the ACO had redesigned the hardware.
> You raised the subject by comparing internals of both calcs.
Do you mean hardware? I don't recall bringing that up.
> Read for example Davenport-Siret-Tournier.
Thanks, I don't have that. You are not suggesting that is a requirement for
using SBS in the '49?
> Which shows that the way we teach in France is somewhat different
> from elsewhere.
Of course. So the SBS is largely only practical for French students.
> French is not as universal as English, but it is also an
> universal language as are e.g. Chinese, Spanish and many other
> languages. It would be a big loss if everything would be replaced
> by one language. And there is no reason that English will remain
> dominant in the future.
Are you expecting English to vanish in the next 20 or 30 years - the life
expectancy of the '49? That seems pretty thin in my eyes. Would you find a
calc using Chinese language a fitting product for the target group of the
'49?
> It seems we are engaged in an endless debate because I don't believe
> the world should be built following the model of the Cathedral
> but instead of the Bazaar.
There is one difference - I used facts, not opinions.
> > You have had 3 and soon 4 years in total. In that time, you could do
> > practically everything, especially with that huge HP financial backing
you
> > had.....
> You can't speak from this subject since you certainly do not have any
> information.
I'm only using the information JYA brought forward. If you say I'm using
false information, look towards JYA, not me. I think it is time for a
decission - did you or did you not have good financial backing from HP?
> Why? When the CAS will be released under the LGPL, people will have
> much more info and tools than we had when we started coding for the 48
> and they will be able to *modify* their OS in flash.
Ok - we'll see when the sources are released. Are the source files improved
over the last year or so? I mean, if they are as messy and badly commented
as a year ago, they are impossibly hard to work with. You can even see how
many problems the ACO have fixing bugs, and they made it. How many new bugs
will be introduced when a "novice" touches the source?
> And
> there are certainly good coders in the young generation. Maybe they
> will prefer coding other projects, maybe not, who knows?
Time will tell. Remember that when people coded for the '48 series, there
were big motivations; A truly superior product and an admirable company.
Both are gone now.
Regards
Steen
Jean-Yves Avenard
"Steen Schmidt" <ssch...@nospam.dk> wrote in message
news:8vPj8.309$ZI6....@news000.worldonline.dk...
> I'm not accusing you for not writing it in English, I'm just saying that
> having it in French is not the big advantage you say it is.
>
Well, as a matter of fact, France is a bigger market than the US regarding
CAS enabled calculator. That's also the reason why the TI92 was first
released in France almost 1 year before any other country.
There have never been a big call for CAS calculator in the US and that's the
reason why the HP40 hasn't been released in this country. As a matter of
fact, most teachers did NOT want HP to sell these calculators in the US.
So French is as good as English on this matter.
> > Why should the root langage always
> > be English?
>
> Because so many uses it?
So what about mandarin ?
> > You can't speak from this subject since you certainly do not have any
> > information.
>
> I'm only using the information JYA brought forward. If you say I'm using
> false information, look towards JYA, not me. I think it is time for a
> decission - did you or did you not have good financial backing from HP?
What are you trying to suggest here ? That I gave you insight on Bernard's
licensing contract ? I dare you!
> Ok - we'll see when the sources are released. Are the source files
improved
> over the last year or so? I mean, if they are as messy and badly commented
> as a year ago, they are impossibly hard to work with. You can even see how
> many problems the ACO have fixing bugs, and they made it. How many new
bugs
> will be introduced when a "novice" touches the source?
I don't recall you being part of the source code distribution list....
So I wonder from which source code you're talking about.
> Time will tell. Remember that when people coded for the '48 series, there
> were big motivations; A truly superior product and an admirable company.
> Both are gone now.
In which way?
100% of the HP48G software is in the HP49. So function-wise they are
similar. Everything you are doing on the HP48 will work the same way on the
HP49 now (if correct flags have been set-up)
It does less.
> Regards
> Steen
Cheers
Jean-Yves
parisse
>
> > I don't control what HP does. I could only give them advices.
>
> But you argued that the beta ROM issue was not a problem, and I argue it is.
> You say that it is not, since everyone reading here will know to use
> v1.19.6. Then I say that almost none of the users are reading c.s.h. Do you
> admit that the beta ROM is indeed an issue now?
>
It would be an issue for people who have more confidence in companies
than individuals. But I believe that in fact the main issue is not
choosing the ROM but doing the upgrade.
> > Why should the root langage always
> > be English?
>
> Because so many uses it?
>
> > Don't you think it is as easy to translate from French
> > than from English?
>
> Not for me, and not for the biggest part of users.
>
But for a corporation it should be as easy. And in fact
that's exactly what was done for the HP40 CAS documentation.
> > You raised the subject by comparing internals of both calcs.
>
> Do you mean hardware? I don't recall bringing that up.
>
No, you were speaking on algorithms details.
> > Read for example Davenport-Siret-Tournier.
>
> Thanks, I don't have that. You are not suggesting that is a requirement for
> using SBS in the '49?
>
No, but I hope that a curious user would try to find info
on the algorithm.
> > Which shows that the way we teach in France is somewhat different
> > from elsewhere.
>
> Of course. So the SBS is largely only practical for French students.
>
I would say that for the euclidean division of integers, not for
the rest.
> > French is not as universal as English, but it is also an
> > universal language as are e.g. Chinese, Spanish and many other
> > languages. It would be a big loss if everything would be replaced
> > by one language. And there is no reason that English will remain
> > dominant in the future.
>
> Are you expecting English to vanish in the next 20 or 30 years - the life
> expectancy of the '49? That seems pretty thin in my eyes. Would you find a
> calc using Chinese language a fitting product for the target group of the
> '49?
>
I was speaking in general.
>
> I'm only using the information JYA brought forward. If you say I'm using
> false information, look towards JYA, not me. I think it is time for a
> decission - did you or did you not have good financial backing from HP?
>
I can't answer to these type of questions.
> Ok - we'll see when the sources are released. Are the source files improved
> over the last year or so? I mean, if they are as messy and badly commented
> as a year ago, they are impossibly hard to work with. You can even see how
> many problems the ACO have fixing bugs, and they made it. How many new bugs
> will be introduced when a "novice" touches the source?
>
Some parts are well commented, some not (and some French comments remain
in the source). Exactly as for any large project.
> > And
> > there are certainly good coders in the young generation. Maybe they
> > will prefer coding other projects, maybe not, who knows?
>
> Time will tell. Remember that when people coded for the '48 series, there
> were big motivations; A truly superior product and an admirable company.
> Both are gone now.
We'll see. I don't think we will have next year affordable calc using
new generation processors. Hence good coders with few money
might find interesting to code new CAS commands and integrate them
in the 49 ROM. And programming the 49 will remain an excellent way to
learn
programming.
Steen Schmidt
> CAS enabled calculator.
Ok.
> What are you trying to suggest here ? That I gave you insight on Bernard's
> licensing contract ? I dare you!
Of course not! I'm simply stating what you have said yourself, and I'm being
told that I do not know what I'm talking about. Then, is that because you
didn't tell the truth? I don't know, and I don't care, really. Who knows the
truth anyway?
> I don't recall you being part of the source code distribution list....
And that means that I have not seen any of the source files? LOL. Are you
calling me a liar?
> So I wonder from which source code you're talking about.
Even though Cyrille didn't want to touch on the subject of the '49 in
Cupertino, do you believe that there were noone there interested in the '49?
I saw some of the sources in California, and I have been sent snippets of it
now and again. I have seen the actual files, and I have quotes from numerous
people, official as well as unofficial, how they are to work with. I know
Werner couldn't even compile the ones you sent him. If you look in your own
sent items folder from Australia, I believe you'll find several quotes to me
about the files.
I have, as many others here, browsed through much of the ROM, so I have also
got a very good idea about the general structure of it.
I have used a great deal of the internal entry points (most CAS) from the
get go, and I have personally contributed with a very large number of bug
reports. Most of this is not publicly known, but *you* know it.
Are you telling me that it is easy to fix the ROM?
> 100% of the HP48G software is in the HP49. So function-wise they are
> similar.
When you get to SysRPL/ML levels, almost nothing is the same. If you stay
with UserRPL, of course most is the same, but that's not what we're talking
about, when talking about changing the ROM, right?
> Everything you are doing on the HP48 will work the same way on the
> HP49 now (if correct flags have been set-up)
Not when it comes to the 1 MB ROM.
Jean-Yves, this is not a war. I'm not battling you or Bernard - I have great
respect for you, but I think I'm entitled to my own opinion about how easy
or hard I think it'll be to fix the ROM by this community.
I don't care if it was the fault of HP or of the ACO, but I believe it's
correct that the documentation was not good at release. I believe that
neither of the release ROMs are good - v1.18 is not stable either. It has
numerous flaws, many may not be obvious to all users, but some are
definetely evident in everyday use. I do believe that not having a stable
release ROM is a problem, as I do not think that as much as 10% of the users
are confident in loading a beta ROM.
These are opinions, most backed by solid arguments - they are not personal
attacks on either of you.
Regards
Steen
Tom Garcia
> Well, we share many points of view, but I'm trying to look at the HP49G as
Yes, I agree on most of the points you just gave. Basically, neither
of us believed what HP claimed about their product, which is a fair
product, but not for the reasons HP claims. A few remaining thoughts
below.
> Had it been a more important application, you wouldn't have dared to use a
> beta (I hope). If not for anything else, then because of the risk of loosing
> the ability to return to the manufacturer with any legal claims.
I think the NO WARRANTY clause on almost every piece of software means
that, like it or not, in most countries there is no chance whatsoever
of making legal claims in the event of damages, whatever the release
status. Again, I'd rather rely on the word of other experts, and my
own judgement.
> point. If you don't miss it, great, but if you do (because you are used to
> HP thoroughnes) then too bad.
I admit, I have read through some of a HP48 manual, and it's a lot
better (when I asked HP, they actually told me to get an HP48 manual
for the 49).
> LOL - have you seen the sources? Good luck, I'm telling you. Good luck.
I've worked with horrible code, and with beautiful code. If it is
accompanied by documentation with info on algorithms etc, it could
prove very interesting, even if just as a basis for further projects,
and/or just to fix the odd minor issue. The ideas and techniques are
more important than the particular implementation.
> > No, students who weren't willing to put in extra effort should have
> > realised that they shouldn't be buying a complex calculator *at all*.
>
> Hmm, the initial target group is usually very subsceptible to commercials.
> They didn't have a chance in my opinion.
There were commercials for the HP49? :-) In England I could only find
about 3 places that sold TI-89's and they were mail order, and 1 place
that sold HP49s (ignoring HP/TI themselves). *Neither* company seems
to have ever made much effort in recent years to push its high end
calculators here. The less powerful TIs seem to be available in the
high street, though.
> You'll just have to wait and see then. I doubt the '49 OS will be the next
> Linux. Only a selcet group will ever grasp to modify the ROM. Werner is one
As above, it's more likely people will take ideas from it for new
projects. Which is good. There's not much use for the code itself in a
non-portable language.
> > I didn't realise the Linux kernel slowed down the average app 50-100
> > times :-).
>
> Ok - I meant percent. Sorry. I surely wouldn't choose Linux if the only
> thing I gained was half the performance.
I'd argue that Linux kernel housekeeping isn't that inefficient :-).
> > Seriously, have you tried running Unix variants on ARMs?
>
> Yes. They are not entirely bad, but in the short time we used Linux, we (a
> body of mine really, who is in charge of the OS) found too many anomalies if
I think "anomalies" is a good way of describing Unix issues. Nothing
quite fits together perfectly. I'd probably argue that the effort to
make it do so is less than simply re-writing *and maintaining* an OS
from scratch -- such projects as BeOS never obtained sufficient apps
software, device drivers, credibility (telnet to a root shell --
oops!), etc.
> The sources are free, of course, but they are not that useful without
> support. We can of course compile them, but what happens when we discover a
> bug? We can't force anyone to fix it. That uncertainty is hard to swallow.
Even with moderate $, you can employ a Linux kernel expert to make
your fixes as and when you need 'em. And you can be pretty sure that
major bugs will be fixed pretty quickly anyway by the standard
contributors. Meanwhile, you usually need HUGE $ to get big OS
manufacturers to fix their software :-).
Good luck with your project,
-- Tom
James M. Prange
> > Most act that way. You'd need very good inside knowledge to know if a given
> > beta was stable or not. There is a reason it haven't got released, and
> > that's because it haven't passed proper testing yet. That measn that it's
> > entirely guesswork if you mean that v1.1.9.6 is stable.
> >
>
> I don't think anyone reading comp.sys.hp48 would doubt that 1.19.6
> is more stable than 1.18. You don't need to be an expert, it's just
> whether you have more confidence in all the users in this forum
> (including the developpers) or in the people who decide to label
> a RO
> M.
What fraction of 49G purchasers have ever read comp.sys.hp48?
The User's Guide that came with my calculator mentions
www.hp.com/calculators/hp49/, which, unless I'm mistaken, did in
the past offer an "unsupported" beta ROM. I don't think you can
find a beta ROM on the hp site now. I don't know about you, but
I generally avoid using software labeled "beta" unless I have a
pressing need for the new features and find the risk worthwhile,
or I just want to play around with it. My first assumption is
that the developers don't have confidence that the product is
fit for release if they still have a beta label on it.
Personally, I still don't think that the 49G is suitable for
"real work" yet.
I think that most calculator purchasers expect to be able to
simply put the batteries in and have the calculator work without
installing any system updates. Does every calculator purchaser
have access to a PC that he can connect to his calculator for
ROM updates? And does he want to go to the trouble of calling HP
for a ROM upgrade kit? Does he realize that that's an option or
does he think that he has to shell out more money for a
"connectivity kit" before he can update the ROM? I wonder how
many have upgraded their 49G at all.
> > Does everybody do that? I do not hope that you are responsible for any
> > mission critical equipment, gievn your lack of concern about using beta
> > software.
> >
>
> I generally tend to follow the developper advice when using their
> software.
Wouldn't most purchasers expect that the developer was HP?
> >
> > So, can you tell me the how SOLVE works, and what its limitations are?
> > What's the maximum order you can develop a series to? Why? Why doesn't
> > PSI(1,n), where n is equal, get evaluated in exact mode? How come GAMMA(1/2)
> > gets evaluated in exact mode, but GAMMA(1/4) doesn't? Why does 'LN(X)' 'X=2'
> > SUBST demand RAD mode?
> >
>
> The kind of documentation you would require would cover 2 books
> of the size of the 48 AUG. I don't think it would have been worth
> the effort (and we would anyway not have time to do it).
Actually, the one for the 48G series is called a "Reference
Manual" not a "Guide". But so what if it would take 2 books that
size? I'd like to have that kind of documentation available from
HP. I wouldn't expect the developers to write the user
documentation; actually they may well have a difficult time
judging the expertise of the users, and I wouldn't demand that
they be highly skilled at writing. I would expect the developers
to write a "rough draft", turn it over to professional writers,
be available to answer questions, and check the professionally
written copy for errors (not spelling errors and such, but
errors about how to use the calculator). And I think that the
full Advanced User's Guide should've been available in printed
form, at least as an optional additional purchase for the sake
of those who don't have a PC and an internet connection. For
that matter, I don't want to have to go to a PC every time I
want information about the calculator, so I printed out a copy
to put in a 3-ring binder. I would've been willing to pay a
reasonable amount to save the time and cost of doing my own
printing. I'm not saying that the programmers were personally at
fault here, but rather that HP (like many other companies) was
very much at fault. I expected better from HP. If they want to
be just a printer and PC clone company, they may as well have
not even bothered with the 49G.
> Does it
> mean we should not have programmed the feature?
Well, how much good is it to have a feature that the users don't
know how to use? Perhaps the programmer gets a sense of
accomplishment, but for the user, obscure undocumented
"features" may very well do more harm than good, and be
perceived as "bugs" rather than "features".
> Should I have
> stopped all work on the 49 in 1999 just because HP would not release
> any stable ROM release?
> Usual features are documented either in HP documentation
> or in Renee's documentation (which was available in French from
> the beginning and was distributed by HP France in 1999)
In French!? Well, no doubt French is a fine language for those
who understand it, but I would expect the documentation to match
whichever language the package is printed in. Couldn't HP hire
translators?
> > LOL - when have you last seen JYA or BP give insights to the '49? HP is just
> > such a company.
> >
>
> As Jean-Yves said, the CAS will be released under the LGPL license.
> That means that starting mid-2003, you will have full access
> to the CAS source and you will be able to make your own CAS. I don't
> think you will ever be allowed to read TI source code.
I personally don't give a rat's ass about TI's source code or
about HP's practices being better than or at least no worse than
TI's. I bought an HP, not a TI.
I don't think that anyone realistically expects such a complex
device to be perfect. But have you ever heard such sayings as
"Anything worth doing is worth doing well." and "Don't bite off
more than you can chew."?
--
Sincerely,
James
Parisse Bernard
I believe people have some legitimate reasons to ask for
more documentation, more support from HP and so on. Yes, it
would have been better if the French doc from Renee had been
translated. I agree HP should relabel 1.19-6 (or even better
the unpublished 1.19-7) as 1.20. Hp could even provide
an upgrade program in some reseller stores so that novice
users can get a stable ROM.
But this is the wrong place to say these things. It's easy
to complain here, but I don't believe HP cares a lot about
this newsgroup. Jean-Yves or Gerald or Cyrille or I, we read
this newsgroup, and we feel bad when we read comments about
the buggy ROM or lack or documentation and so on because we fixed bugs
or we provided documentation. Unfortunately we are not in a situation
to make it "official". Just imagine you would be
in our situation, how would you react reading these type of
posts? The result of complains here is probably the exact
opposite of what people would like to have: e.g. I could decide to
ignore comp.sys.hp48 in the future, or even I could decide
that releasing the CAS under the LGPL is not worth the effort...
Therefore I suggest that you find a more productive way to
complain: for example, I believe that sending a snailmail
explaining these issues to HP could have more impact.
R Lion
I'm user of your Erable3.201 beta in my 48GX. Also user of MetaKernel 2.3
I only want to thank you two, your work for developing these programs years
ago: they make my diary work more comfortable.
Best Regards
Raul Lion
(By the way: Look my surname. I'm spanish, but the father of my granfather
was a french engineer)
"Parisse Bernard" <par...@fourier.ujf-grenoble.fr> escribió en el mensaje
news:3C909958...@fourier.ujf-grenoble.fr...
Tom Garcia
> this newsgroup. Jean-Yves or Gerald or Cyrille or I, we read
> this newsgroup, and we feel bad when we read comments about
> the buggy ROM or lack or documentation and so on because we fixed bugs
> or we provided documentation.
My participation in this thread has been to defend the quality of the
ROM, and to highlight that the problems described are with HP
marketing, and not the result of your excellent efforts.
Do not be disheartened when people criticise. This means people feel
passionately about your project. Plus, many have come to your defence,
in this and other threads. I don't often speak up, but I felt you
deserved this defence.
I look forward to the release of the ROMs under (L)GPL, so I can
investigate them for my own interest, and contribute if I have time.
Thanks for putting in your best efforts. They have not gone unnoticed.
-- Tom
R Lion
"R Lion" <pit...@NOyahooSPAM.es> escribió en el mensaje
news:a6qpla$gj9$1...@titan.bt.es...
> This is not poetry:
> I'm user of your Erable3.201 beta in my 48GX. Also user of MetaKernel 2.3
> I only want to thank you two, your work for developing these programs
years
> ago: they make my diary work more comfortable.
> Best Regards
> Raul Lion
>
> (By the way: Look my surname. I'm spanish, but the father of my granfather
> was a french engineer)
>
...and just in case: what I say above, is against NOBODY.
This group works
Raul
Steen Schmidt
> more documentation, more support from HP and so on. Yes, it
> would have been better if the French doc from Renee had been
> translated.
That's what I've been saying all the time.
> I agree HP should relabel 1.19-6 (or even better
> the unpublished 1.19-7) as 1.20.
That too.
> Hp could even provide
> an upgrade program in some reseller stores so that novice
> users can get a stable ROM.
That would be great.
> But this is the wrong place to say these things.
Why? I wasn't complaining. This thread started as a discussion about how
hard or easy it would be for this group to continue the work of You and the
ACO. I provided my opinion about the state of the '49 - the need for a
stable release ROM, the lack of English documentation and so on.
If I must be so free, it is You and JYA who have continued on about there
was nothing wrong. You have been contributing very much to the length of
this thread. If this last post of yours is a benchmark of your opinion, we
share the same.
I never said this was your fault. I never blaimed the ACO for this situation
we're in now. I have always stated that I believe it is HPs bad, so there
was never any accusation against any of you.
The reason I will not continue any work on the '49, is because I won't help
HP sell even a pencil from now on. They are acting so bad, that I will not
ever again be on their side. They are not on mine.
> It's easy
> to complain here, but I don't believe HP cares a lot about
> this newsgroup.
My statements here were never to HP.
> Jean-Yves or Gerald or Cyrille or I, we read
> this newsgroup, and we feel bad when we read comments about
> the buggy ROM or lack or documentation and so on because we fixed bugs
> or we provided documentation. Unfortunately we are not in a situation
> to make it "official". Just imagine you would be
> in our situation, how would you react reading these type of
> posts?
Is that why you're defending HP then? You say: "The beta label on the stable
ROMs is not a problem", "The documentation is there, but it is in French -
that is not a problem though".
I understand you feel hurt, but it's HP you should look towards, not us.
We're not complaining about your effort - on the contrary, I believe most
here think of you as heroes. You were once the small men, part of a group
that made software for the HP48G series. Now you have made it with the big
company, HP, but they unfortunately dropped the ball and stabbed you in the
back.
That's why I said that I wouldn't have continued my work if I'd found out
what HPs plans were. It's a big insult towards you, when HP sacks the ACO,
and disses the HP49G the way they have.
You must feel 100-fold the disappointment I do, so you must be pretty
pissed.
> The result of complains here is probably the exact
> opposite of what people would like to have: e.g. I could decide to
> ignore comp.sys.hp48 in the future, or even I could decide
> that releasing the CAS under the LGPL is not worth the effort...
> Therefore I suggest that you find a more productive way to
> complain: for example, I believe that sending a snailmail
> explaining these issues to HP could have more impact.
These are not complaints. This thread really started when George Tsiros
called Micah a troll, because he thought it was futile to rewrite the ROM. I
gave my opinion of the current state of the '49, and you and JYA did nothing
but defend the '49 and its development.
I understand you feel hurt, but nobody said it was your fault - everyone
knows it is the fault of HP. If you defend their decissions, it could look
like you thought that HP handled this correctly. I'm glad you finally don't.
Regards
Steen
MSCHAEF.COM
Tom Garcia <tga...@hivemind.org> wrote:
>
>I look forward to the release of the ROMs under (L)GPL, so I can
>investigate them for my own interest, and contribute if I have time.
I've seen this mentioned a couple times before, with varying details,
and had a couple questions.
First, is the release being done under LGPL or GPL?
Second, does it encompass just the CAS, or the entire firmware source
code? I assume that if it's GPL, the 'viral' nature of that license will
require it to be the entire firmware.
Thanks for the info. Also, to the authors, thanks for all the effort
you've put into the HP. After the GX, I had given up the 48 series for
dead, and am grateful that the 49 exists at all.
-Mike
James M. Prange
> I begin to think that this thread should stop, here is why:
> I believe people have some legitimate reasons to ask for
> more documentation, more support from HP and so on. Yes, it
> would have been better if the French doc from Renee had been
> translated. I agree HP should relabel 1.19-6 (or even better
> the unpublished 1.19-7) as 1.20. Hp could even provide
> an upgrade program in some reseller stores so that novice
> users can get a stable ROM.
> But this is the wrong place to say these things. It's easy
> to complain here, but I don't believe HP cares a lot about
> this newsgroup.
Actually, I get the impression that HP doesn't care a lot about
calculators or calculators purchasers anymore.
> Jean-Yves or Gerald or Cyrille or I, we read
> this newsgroup, and we feel bad when we read comments about
> the buggy ROM or lack or documentation and so on because we fixed bugs
> or we provided documentation. Unfortunately we are not in a situation
> to make it "official". Just imagine you would be
> in our situation, how would you react reading these type of
> posts?
Ok, I came down on you harder than I should have. I apologize for that;
it's just that you touched on some of my pet peeves about the 49G, and I
got the impression that you were saying that they really weren't
significant problems. I do think that we can all at least agree that the
49G is not as good as it should've been.
And yes, new ROMs were released, with substantial improvements and bug
fixes, apparently from you guys continuing development on your own time,
after, as far as I can tell, HP stopped supporting continued
development. And you've made a good deal of "unofficial" documentation
available if one knows where to search, even if it's not in
professionally written, printed and bound manuals. Most people would've
dropped the effort when the company stopped supporting it, but you guys
went beyond what was required by HP. And even now you still show an
interest in this newsgroup and the 49G. You have my sincere thanks for
all of that.
I expect that the four of you feel even worse about the way things have
turned out than the rest of us do. I don't place the blame on the
developers personally. I know all too well how badly corporate
management can damage or destroy a promising project with changing
requirements, "hurry up", "that's good enough, we're not building a
space shuttle you know", basically "drop whatever you're working on and
do this instead", and generally demanding results that make them look
good right now; why should they care about the future of the group? They
plan to have moved on to better things by then. And so on. And marketing
really blew it; I might never have realized that the 49G even existed if
I hadn't happen to notice the HP logo on a weird looking calculator
behind a dusty 48GX at Office Depot. That's mostly speculation of
course; I have no way of knowing for sure just what went wrong, but I
expect that the blame really belongs rather higher up the corporate
ladder (and yes, I realize that you personally weren't even on that
ladder at all).
> The result of complains here is probably the exact
> opposite of what people would like to have: e.g. I could decide to
> ignore comp.sys.hp48 in the future, or even I could decide
> that releasing the CAS under the LGPL is not worth the effort...
I sincerely hope that some form of development can be picked up sometime
in the future. I don't know that there's any great need for new
features, (not all will agree with that, I know) but certainly I'd like
to see the known problems fixed to whatever extent is feasible. And I
hope that the documentation becomes clearly "open source" as well.
I also wish Steen the best with his plans. Perhaps someday he'll produce
the calculator that HP should've made. HP seems to have dropped
development yet again, and who knows whether they'll ever even try to
pick up the pieces again.
> Therefore I suggest that you find a more productive way to
> complain: for example, I believe that sending a snailmail
> explaining these issues to HP could have more impact.
I suppose miracles might happen. Whom can we write to who might pay
attention to what we say and might be in a position to change things?
--
Regards,
James
Detlef Mueller
> In article <616e1c01.02031...@posting.google.com>,
> Tom Garcia <tga...@hivemind.org> wrote:
>
>>I look forward to the release of the ROMs under (L)GPL, so I can
>>investigate them for my own interest, and contribute if I have time.
>
> I've seen this mentioned a couple times before, with varying details,
> and had a couple questions.
>
> First, is the release being done under LGPL or GPL?
According to the 1st post from JYA mentioning it it'll be LGPL
(http://www.gnu.org/copyleft/lesser.html)
> Second, does it encompass just the CAS, or the entire firmware source
> code? I assume that if it's GPL, the 'viral' nature of that license will
> require it to be the entire firmware.
Sounds like just the CAS and MK to me, the rest is probably HP property...
Bye,
--
`What a depressingly stupid machine' Detlef Mueller
-- Marvin Detlef[DOT]M[AT]hamburg[DOT]de
http://mein.hamburg.de/homepage/grendel
not waving but drowning
>I would say any bright high school kid or (university) >college student in
any of the hard sciences/engrg, maths >or computing should be able to grasp
an HP. That's the >level I was thinking of, compared to other makes which
>anyone from high school onwards should be able to >grasp quite easily.
I know kids who are using their HP 48 in ninth grade.
"Tom Garcia" <tga...@hivemind.org> wrote in message
news:616e1c01.02031...@posting.google.com...
G Savage
"Jean-Yves Avenard" <jean...@avenard.org>
> Well, as a matter of fact, France is a bigger market than the US regarding
> CAS enabled calculator. That's also the reason why the TI92 was first
> released in France almost 1 year before any other country.
> There have never been a big call for CAS calculator in the US and that's
the
> reason why the HP40 hasn't been released in this country. As a matter of
> fact, most teachers did NOT want HP to sell these calculators in the US.
If this is really so, I think it's a very interesting fact and am curious to
know why. Does anybody have explanations? How are French or American
students benefiting, or not benefiting, from CAS and the like? I am
involved with my daughter's school and am trying to help raise the standards
in math as well as developing a proper approach for the use of computers in
learning.
Greg S
Raymond Hellstern
you forgot to snip the remaining thousand lines below your 1-line comment;-)
Or is it another pseudonym for VPN? :-))
Raymond
"not waving but drowning" <pace...@hotmail.com> schrieb im Newsbeitrag
news:5idk8.59338$Yd.37...@e3500-atl1.usenetserver.com...
Werner Huysegoms
> > So I wonder from which source code you're talking about.
>
> Even though Cyrille didn't want to touch on the subject of the '49 in
> Cupertino, do you believe that there were noone there interested in the '49?
> I saw some of the sources in California, and I have been sent snippets of it
> now and again. I have seen the actual files, and I have quotes from numerous
> people, official as well as unofficial, how they are to work with. I know
> Werner couldn't even compile the ones you sent him. If you look in your own
> sent items folder from Australia, I believe you'll find several quotes to me
> about the files.
>
I have never sent anyone any part of the source code I received. (I know
you didn't say that I did - it's just for the record)
The reason that I could not compile it is that I couldn't get the Cygwin
tools to function on my Win98 home machine - something Jean-Yves warned me
about in advance - but I could hardly install them on the WinNT at work..
Good luck with the Steen-machine, Steen!
Werner
R Lion
"not waving but drowning" <pace...@hotmail.com> escribió en el mensaje
news:5idk8.59338$Yd.37...@e3500-atl1.usenetserver.com...
> Snip
> >I would say any bright high school kid or (university) >college student
in
> any of the hard sciences/engrg, maths >or computing should be able to
grasp
> an HP. That's the >level I was thinking of, compared to other makes which
> >anyone from high school onwards should be able to >grasp quite easily.
>
> I know kids who are using their HP 48 in ninth grade.
>
How old are they?
Thanks
Detlef Mueller
Why not using the F21 (http://www.ultratechnology.com/f21.html) ?
Everything you need, the existing HP code should be easy to port
and the thing can run @ 2700 MIPS. See also http://www.ultratechnology.com/chips.htm
Steen Schmidt
> Everything you need, the existing HP code should be easy to port
> and the thing can run @ 2700 MIPS. See also
http://www.ultratechnology.com/chips.htm
Are you kidding? The F21 supports 15 colors at some unknown RGB resolution,
it'll address 5 Mbyte DRAM, 1 Mbyte Flash and 80 kbyte SRAM. It's
coprocessor runs at 40 MHz, and does not really have that many features. In
comparison, the PXA250 supports a video resolution of 800x600 @ 64k colors,
256 Mbyte DRAM, something like 64 Mbyte SRAM and 256 Mbyte Flash. It's
coprocessor contains a DSP and runs at speeds up to and exceeding 1 GHz -
the same with the processor core.
The F21 is a toy compared to the XScale implementations, but it would fit
nicely in the Saturn spirit. Is there something I misunderstand?
Regards
Steen
rcobo
Cyrille de Brébisson
Well, personnaly, I would agree to go Xscale, it's at this point in time the
best bang for the money and power. it's a real CPU, with MMU, 2*16K cache.
True, there is not FPU, but quite frankly, how often do you perform real
number crunching on your calculator?
regards, Cyrille
"Steen Schmidt" <ssch...@nospam.dk> wrote in message
news:oJNl8.8962$FT.3...@news010.worldonline.dk...
Veli-Pekka Nousiainen
news:t7Xj8.11226$uR5....@newsfeeds.bigpond.com...
> Hello
>
> "Steen Schmidt" <ssch...@nospam.dk> wrote in message
> news:8vPj8.309$ZI6....@news000.worldonline.dk...
> So French is as good as English on this matter.
The first (foreign) languge in Europe is typically
English, Germany, Spanish, French, Russia?
Th company language for almost all global companies
is - you guessed correctly - English
X
> 100% of the HP48G software is in the HP49. So function-wise they are
> similar. Everything you are doing on the HP48 will work the same way on
the
> HP49 now (if correct flags have been set-up)
X
I still can't find everything...
For example making own functions in UserRPL
while using APPLY doesn't give me a change
to evaluate variables stepwise...
Where are the equations?
TEACH & CLTEACH ? (which could be revised)
How about 80% instead of 100%
VPN
PS: Thank you a lot of your good work!
PPS: Any guesses of the 1.19-7 release date?
Veli-Pekka Nousiainen
http://www.mips.com/publications/documentation/MD00207-2B-20KCTSMC18-DTS-01.
07.pdf page 13 virtual page numbers OR
http://www.mips.com/publications/documentation/MD00014-2B-5KC-DTS-02.23.pdf
page 11 virtual page numbers...;-)
VPNnews:oJNl8.8962$FT.3...@news010.worldonline.dk...
Veli-Pekka Nousiainen
VPN
"Raymond Hellstern" <Raymond....@T-Online.de> wrote in message
news:a6s6qh$tth$07$1...@news.t-online.com...
> Hi,
>
> you forgot to snip the remaining thousand lines below your 1-line
comment;-)
> Or is it another pseudonym for VPN? :-))
>
> Raymond
>
> "not waving but drowning" <pace...@hotmail.com> schrieb im Newsbeitrag
> news:5idk8.59338$Yd.37...@e3500-atl1.usenetserver.com...
> > Snip
Veli-Pekka Nousiainen
news:a6lirn$87r$1...@web1.cup.hp.com...
X
> > Perhaps even you would have had enough time and money to design a
> > faster C.P.U and a better controller than the Yorke.
> Well, ASIC design time is in years from starting the work to first
> prototypes. the 49 was introduced barely a year after the ACO was created,
> Even with an experienced team, it would not have been possible to redesign
> the Yorke in such a small timeframe.
X
BUT how about just downscaling it from 1 micron to 0.25 micron
gaining 1:16 in yield of chips and achieving easily 20MHz
while reducing power consumption all at the same time?
The ignorant anti-rcobo
Veli-Pekka Nousiainen
> Hello,
>
> Well, let us put it that way, the saturn delivers roughly 1.6 Mips / What,
a
> Intel Strong arm 1100 at 133Mhz delivers 200 Mips / What
> This means that if the SATURN (or Yorke) was redesigned now, it would be
125
> * faster for the same power usage (or 125 times less power hungry at the
> same speed)....
> The saturn uses roughly 20mA at 5V, this means 100mW, as a comparison, an
> Intel PXA250 at 300Mhz uses 300mW (and, as power consumption is factor of
> the square of the speed, this means that the same CPU should disipate the
> same power than the saturn at 173Mhz!)!
>
> Now, truth is, performance per mips is only an issue in some embedded
> systems, so it is difficult to find a low computing power / low power
usage
> CPU, but have you look at embedded MIPS or embedded Arm7 platform? you
will
> find that the Saturn is not that good!
>
> Conservatives estimates 3 years ago showed that the Saturn could be
> redesigned to run at 40Mhz with a lower power consuption...
>
> Regards,Cyrille
X
That is exactly what I was hoping for !!
downscaling the pine-sized lines to a reasonable 0.25 micron
should give that performance boost to the HP 49G
This could be done NOW !!!
VPN
Jean-Yves Avenard
"Veli-Pekka Nousiainen" <DROP...@welho.com> wrote in message
news:a7sg5t$6ic$1...@tron.sci.fi...
> The first (foreign) languge in Europe is typically
> English, Germany, Spanish, French, Russia?
> Th company language for almost all global companies
> is - you guessed correctly - English
> X
I doubt that English is more spoken in Europe than any other language.
Spanish is spoken only in Spain (and not even in the entire spain).
English is spoken only in the UK and Ireland
German is spoken in Germany, Austria and Switzerland
French is spoken in Belgium, France, Luxemburg and switzerland
I guess it's a very hard match between French and German in Europe...
Jean-Yves
R Lion
"Jean-Yves Avenard" <jean...@avenard.org> escribió en el mensaje
news:omno8.23620$uR5....@newsfeeds.bigpond.com...
> Spanish is spoken only in Spain (and not even in the entire spain).
> English is spoken only in the UK and Ireland
Not very exact: Spanish is spoken in (entire) Spain, and almost other
lenguages.
Exactly as English is spoken in UK and Ireland and almost Gaelic, and
perhaps others I don't know.
Raul
Jean-Yves Avenard
Some people in Barcelona's surrounding would argue that :)
Same about the Basque country
Jean-Yves
"R Lion" <pit...@NOyahooSPAM.es> wrote in message
news:a7t0gh$kaj$1...@titan.bt.es...
Jordi Hidalgo
> Some people in Barcelona's surrounding would argue that :)
> Same about the Basque country
Unfortunately we are coerced to learn the Kingdom's language; but
nothing (not even 40 years of fascism) can prevent Catalans from
speaking our beloved language, which is spoken in four countries:
Andorra, Spain, France and Italy (Alguer).
The case of the Basque language is worse, because Basque speakers
are seen as separatists (and therefore as terrorists) in Madrid.
We need Peace, Republic and more HP calcs!
Regards,
Bye.
Jordi Hidalgo
jo...@tv3mail.com
-
--
Posted via Mailgate.ORG Server - http://www.Mailgate.ORG
R Lion
Talking in this way is not separatism?
Raul
"Jordi Hidalgo" <csh...@sputnikmail.com> escribió en el mensaje
news:43327b4bbe2fd2720b6...@mygate.mailgate.org...
R Lion
> Hello
>
> Some people in Barcelona's surrounding would argue that :)
> Same about the Basque country
>
> Jean-Yves
Nobody can argue what I said: "Spanish is spoken in (entire) Spain, and
almost other lenguages."
It is a fact.
Raul
Carlos J. Bourlot
9 countries only in South America where the spanish is the first
language. In Central America happens the same...
Regards,
Carlos J. Bourlot
"Jean-Yves Avenard" <jean...@avenard.org> wrote in message news:<H1vo8.23829$uR5....@newsfeeds.bigpond.com>...
R Lion
news:9f759c42.02032...@posting.google.com...
> I'm from Argentina where ALL speak spanish; and by the way, there are
> 9 countries only in South America where the spanish is the first
> language. In Central America happens the same...
>
> Regards,
>
> Carlos J. Bourlot
>
>
Saludos Carlos:
No creas que olvido eso: este es uno de los idiomas más universales. A pesar
de ello, algunos por aquí, reniegan de él y dicen sentirse ¡coaccionados! a
aprenderlo... 40 años de fascismo que no permitió la enseñanza ni la
difusión de las otras lenguas españolas, han sido seguidos en algunos casos,
de 30 años de políticas educativas igualmente dañinas y equivocadas,
tratando de reducir la presencia y el uso del español en las escuelas y en
el resto de la sociedad. Este es un triste tema que conozco bien, ya que soy
profesor (y vasco en un 50% de mis apellidos)
La razón de que no "se os tuviera en cuenta" en la conversación, es que se
estaba hablando de la lengua adecuada para el mercado Europeo.
Recibe un cordial saludo
Raúl
R Lion
"R Lion" <pit...@NOyahooSPAM.es> escribió en el mensaje
news:a7vhin$4r1$1...@titan.bt.es...
> No creas que olvido eso: este es uno de los idiomas más universales. A
pesar
> de ello, algunos por aquí, reniegan de él y dicen sentirse ¡coaccionados!
a
> aprenderlo... 40 años de fascismo que no permitió la enseñanza ni la
> difusión de las otras lenguas españolas, han sido seguidos en algunos
casos,
> de 30 años de políticas educativas igualmente dañinas y equivocadas,
> tratando de reducir la presencia y el uso del español en las escuelas y en
> el resto de la sociedad. Este es un triste tema que conozco bien, ya que
soy
> profesor (y vasco en un 50% de mis apellidos)
> La razón de que no "se os tuviera en cuenta" en la conversación, es que se
> estaba hablando de la lengua adecuada para el mercado Europeo.
>
> Recibe un cordial saludo
>
> Raúl
>
>
Free translation into (bad) english. I hope you can understand what I wanted
to say Carlos.
I don't forget what you say: this (spanish) is one of most used lenguages in
the world. The second, I think...
In spite of it, some spanish that don't feel spanish, renounce this
lenguage. They say they're coerced! to learning it (and I'm sure that many
of them, pay for english lessons :-)
40 years of fascism with any other spanish lenguage forbidden: very bad. And
now, in some cases, 30 years of local education systems as wrong and
harmfull, trying to eliminate spanish of schools and society. Sadly, I know
it well because education, is my work.
Don't feel forgotten: we were talking about the most apropriate lenguage for
european market.
R Lion
Newsreaders on TV, say "Londres" instead "London". Normal.
They say "Munich" instead "München". Normal.
"Reino Unido", "Irlanda", "Alemania", instead "United Kingdom", "Ireland"
(or Eyre!) or "Deutschland". Normal: they're talking in spanish.
BUT:
they say "Girona" (catalan) instead "Gerona" (spanish).
"Euskadi" (basque) instead "Pais Vasco" (spanish).
"Lendakari" (basque) instead "presidente" (spanish),
as when they're talking about the german, american or wherever president of
the world...
and many many more...
I feel the "lenguages/flags war" is very sad...
Raul
Carlos J. Bourlot
resto de Sud América en gral. pasó lo mismo con las lenguas
aborígenes.
Saludos desde Argentina,
Carlos
<english>
I completly agree with you Raúl!..... In Argentina (and south america
in general happened the same thing with the language of the native
people of this part of the world.
Regards from Argentina,
Carlos
</english>
"R Lion" <pit...@NOyahooSPAM.es> wrote in message news:<a7vmhb$7ij$1...@titan.bt.es>...
HPhreacker 2002(®)
R Lion escreveu na mensagem ...
>This is not poetry:
>I'm user of your Erable3.201 beta in my 48GX. Also user of MetaKernel 2.3
>I only want to thank you two, your work for developing these programs years
>ago: they make my diary work more comfortable.
>Best Regards
>Raul Lion
>
>(By the way: Look my surname. I'm spanish, but the father of my granfather
>was a french engineer)
You're not the unique. I do use Erable, ALG48(Erable has some parts from it!
Trust me!), used ALGB on a HP48G32K, the LIB789 for HP49(some nice commands)
HPhreacker 2002(®)
>> > HP49, I installed 1.19.6. It was completely painless.
>>
>> Were you able to determine in two days, that ROM v1.1.9.6 was stable?
You're
>> one in a million, man.
>
>I was able to judge that, given people's comments, it was worthwhile
>trying it out. I could always go back, and hey, it was a calculator,
>who cares if I get a few hours of poor performance? It was a weekend
>anyway. If the move had been irreversible, I'd have taken longer.
I've installed 1.19-6 in Emu48. Even Emu48 works faster(why? w/ rom 1.18 it
was laggy for "Load object...")
>> > Do you install Microsoft's latest OS the day of its gold release,
>>
>> I can see you're a Linux guy, and colored at that too.
>
>No, I'm mostly a FreeBSD guy and a Windows guy. Whatever gets the job
>done.
>
I have *nothing* against Windows/Microsoft/IE. Imagine if we had only linux
servers: no "Server is busy" Win2000 messages :D.
Veli-Pekka Nousiainen
but, well - simply: you and I do speak English, don't we?!
e.g.. English, French, Spanish, German languages are taught
in school! That was my point.
Now - how many countries in Europe forces students to learn
at least one foreign language - AND
which one is THEN the most common spoken language in Europe?
asks
VPN
"Jean-Yves Avenard" <jean...@avenard.org> wrote in message
news:omno8.23620$uR5....@newsfeeds.bigpond.com...
vIRCiated(tm)
"dan watts" <danie...@hotmail.com> escreveu na mensagem
news:206ab181.02030...@posting.google.com...
> I'm pretty much a newbie. A respected electronics Professor
> recommended the RPN way, and I'm so glad I followed his advice. It
> enabled me to score significantly higher marks than my classmates. In
> studying precalculus I find the Equation Writer and the Step-by-Step
> mode very useful. I love the solid feel of the 49g keyboard. I printed
> and bound Renee De Graeve's "Calcul formel et Mathematiques avec la
> HP49G en mode algebrique", translated into English by Ivan Bertolotti,
> from the website www-fourier.ujf-grenoble.fr/~degraeve/usflan.pdf, and
> have found it a terrific resource. Nick Karagiaouroglou's Trig
> Marathon is a tremendous resource, as will be the upcoming Complex
> Number Marathom and Calculus Olympics. For me, HP is the place to be.
Personally I prefer RPN. Rulez for complex expressions. I found it superior
to Algebraic.
With RPN I don't need Pretty-Print and the like.
> I took a 4-hour electronics exam recently with a public utility. Their
> wording was "a calculator will be the only study aid allowed during
> the test." I hadn't had any exposure to digital circuits, so I had 5
> weeks to learn enough digital to pass the exam. I downloaded Robin
> Getz's Digitician from hpcalc.org website, which takes Karnaugh map
> data in decimal string format and uses the Quine-McCluskey algorthm to
> convert to a simplified Boolean expression. This proved to be very
> useful. I was able to guide my shaky Boolean abilities and check my
> answers with it.
> My digital textbook, "Digital Systems" by Ronald Tocci says, "The
> following discussion (of the Karnaugh map method) will be limited to
> problems with up to four inputs, since even five- and six-input
> problems are too involved and are best done by a computer
> program."(pg. 122) It is interesting to note that Getz's Digitian,
> even while not written in "machine code" and "sacrificing program
> memory and speed for low run time memory", is able to solve a 6-input
> circuit with 15 highs and 10 don't cares in 8 minutes. On the 3- and
> 4-input circuits on the test I took, the solve time seemed plenty
> fast, maybe around a minute. My wife says I love my calc more than I
> love her, and sometimes I'm afraid she's right!
> I have a question for Steen. Do you really think you are going to able
> to bring a superior calc to market any time soon at under $400? Are
> you going to be able to get it on the shelf at the Best Buys and
> Circuit Cities and Office Maxes and their international equivalents
> next to the TIs? Are you going to be able to break TI's stranglehold
> on the schools? I sincerely hope you are successful in doing so. But
> it seems to me the odds of your success are maybe one in a thousand.
> And if you are successful, are you going to give your HP predecessors
> due credit?
W
know you.